Relationships to Reported and Observable Emotional Behavior Through Interpersonal Goals

BELIEFS ABOUT THE SPREAD OF EMOTIONS: RELATIONSHIPS TO REPORTED AND OBSERVABLE EMOTIONAL BEHAVIOR THROUGH INTERPERSONAL GOALS

A Dissertation submitted to the Faculty of the Graduate School of Arts and Sciences of Georgetown University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Psychology

Nicole Senft, M.S.

Washington, DC June 6, 2016

BELIEFS ABOUT THE SPREAD OF EMOTIONS: RELATIONSHIPS TO REPORTED AND OBSERVABLE EMOTIONAL BEHAVIOR THROUGH INTERPERSONAL GOALS

Nicole Senft, M.S.

Thesis Advisor: Yulia Chentsova-Dutton , Ph.D.

ABSTRACT

Beliefs about emotions play an important role in guiding emotional behavior. Though emotions are interpersonal phenomena, we know little about individuals’ beliefs about the consequences of their emotions for others. My dissertation research examines the role of a novel construct—beliefs about the spread of emotions—in guiding emotional behavior, and the influence of self-focused vs. other-focused interpersonal motivations in explaining these relationships. This research was conducted in the U.S. and South Korea because past research has shown other-focused interpersonal motives tend to be more salient in this context than in the

U.S. Studies 1-2 examined the nature of beliefs about the spread of emotions and their relationships to self-reported emotional behaviors. In both the U.S. and South Korea, the more individuals believed their happiness and sadness spread, the more they reported expressing their positive and negative emotions. Interpersonal motivations to strengthen relationships and receive support explained these relationships. Study 3 experimentally manipulated beliefs about the spread of emotions in the U.S. and South Korea. In a simulated social interaction between close friends, I assessed the effects of the manipulation on participants’ efforts to regulate emotions, actual emotional behavior, and expected quality of social support. I also examined the mediating and moderating roles of interpersonal motives in explaining these relationships. Condition did not significantly influence positive emotional behavior, but in both the U.S. and Korea, people led to believe their emotions spread expressed less anger when talking about their negative

iii experiences. This effect was moderated by interpersonal motives. For participants led to believe emotions spread, other-focused motives were associated with increased nonverbal anger expressivity among Americans but decreased expressivity among Koreans. Furthermore, in both the U.S. and Korea, participants led to believe emotions spread expected to receive higher quality support than participants in other conditions. These results suggest that beliefs about the spread of one’s emotions to others predict reported and actual emotional behavior, but in different ways. These findings further suggest that individuals may view the spread of their emotions to others as a useful means of achieving interpersonal goals.

Acknowledgments

This dissertation would never have been written if not for the incredible people that have mentored and supported me throughout the process.

First, thank you to all of the amazing scientists that have invested in my development. To my committee members, I am so grateful for the service you’re doing in sharing your time and expertise. Your insights are so valuable to me. I know I will have a stronger dissertation because of them, and will carry your advice with me through my next steps. A special thanks to Jerry Parrott, it has been such a pleasure to work with you these past few years. Jeanne Tsai, thank you for taking me under your wing and making sure I’ve always had an SPSP family, I can’t tell you how much it meant to me as a nervous, new student who barely knew anyone. José Soto, thank you for being a part of that family—seeing you and your students is among the highlights of every conference. Vivian Dzokoto, thank you for bringing me in on the project that turned into my first peer-reviewed publication, and for helping me to get a job. Mary Murphy, thank you for introducing me to the world of research in social psychology. Your mentorship got me started, you taught me the skills I needed to move forward in the field, and I’m so thankful.

Most importantly, thank you to my advisor, Yulia Chentsova Dutton. You took a risk on a green young girl with slight dependency issues and helped me gain the skills and confidence to be an independent researcher. Your thoughtfulness in research and teacher is inspiring.

To the members of the Culture and Emotions Lab, I am so grateful. Eunsoo Choi, thank you for being so willing to talk through the ideas and methods behind this research, and for being so instrumental in actually conducting it. Thanks are also owed to the many research assistants who contributed to this project at one time or another: Melis, you are my facial coding hero. Zoe Bridges-Curry and Sergio Gaudix, trusting you with data collection was one of the easier decisions I’ve made. Brenda Flores, Sarah Rabon, and Isabelle Ruiz de Luzuriaga, those transcriptions take forever, thank you for shouldering the load with me.

I would like to acknowledge my funding through the National Science Foundation’s East Asia and Pacific Summer Institute. A special thanks is due to Dr. Mark Suh and all the members of his lab at Yonsei University in Seoul for welcoming me into the lab and being incredibly helpful throughout the summer as I worked through the complexities and inevitable difficulties of cross- cultural research. In particular, many thanks to Hwaryung Lee and Jieun Kim for their help with preparing for and carrying out Korean data collection, and to Soyeon Choi and Cheongil Kim for running nearly 100 participants in five weeks from a hallway with no air conditioning. I could not have done it without you.

To all the members of the Georgetown community, I feel so lucky to have been part of a department that supported each other with such genuine enthusiasm.

To Anna, Erika, and Kristin, I am proud to call you my colleagues, and incredibly grateful to call you my friends. When I leave DC, it’s you I’ll miss most of all.

To my family, thank you for making sure I get out of my own head sometimes and for being the people that make me laugh the hardest.

And Jordan, you’re my favorite. Thank you for all of the instrumental support—from teaching me STATA to proofreading job applications. But more importantly, thank you for all the hugs.

Nicole Senft

Table of Contents

Chapter 1: Introduction ...... 1

Chapter 2: Understanding the Nature of Beliefs about the Spread of Emotions and Exploring their Relationships to Reported Emotional Behavior ...... 22

Chapter 3: Explaining Relationships between Beliefs about the Spread of Emotions and Reported Emotional Behavior in the U.S. and South Korea ...... 41

Chapter 4: Manipulating Beliefs about the Spread of Emotions: An Investigation Testing the Causal Relationship between Beliefs and Observable Behavior ...... 72

Chapter 5: General Discussion...... 114

Appendix: Experimental Manipulations ...... 126

References ...... 132

vii

List of Figures

Figure 1: Mean levels of beliefs about the spread of happiness and sadness among Koreans and Americans ...... 69

Figure 2: Mediation model displaying effects of belief in the spread of happiness on reported behaviors through the motive to feel close to others ...... 70

Figure 3: Mediation model displaying effects of belief in the spread of sadness on reported behaviors through the motive to feel close to others ...... 71

Figure 4: Koreans’ and Americans’ mean levels of reported emotion regulation effort in each experimental condition ...... 107

Figure 5: Mean levels of nonverbal expressivity in each condition while discussing negative emotional experiences ...... 108

Figure 6: Koreans’ and Americans’ mean levels of expected support in each experimental condition ...... 109

Figure 7: Relationships between the motivation to benefit others and nonverbal anger expressivity in each condition, separated by culture ...... 110

Figure 8: Relationships between the motivation to strengthen relationships and nonverbal anger expressivity in each condition, separated by culture ...... 111

Figure 9: Relationships between the motivation to receive support and nonverbal anger expressivity in each condition, separated by culture ...... 112

Figure 10: Relationships between the motivation to receive support and expectations of support in each condition ...... 113

viii

List of Tables

Table 1. Exploratory factor analysis of the BASE scale ...... 35

Table 2: Study 1B Demographics ...... 36

Table 3: Correlations between BASES scales and all variables of interest ...... 37

Table 4: Results of hierarchical linear models predicting reported emotional behaviors ...... 38

Table 5: List of items included in the revised BASE scale ...... 59

Table 6: Correlations between all dependent variables among Koreans (lower half) and Americans (top half) ...... 60

Table 7: Results of hierarchical linear models predicting reported emotional behaviors ...... 61

Table 8: Results of hierarchical linear models predicting interpersonal motives ...... 65

Table 9: Indirect effects and bias-corrected confidence intervals of IV on DV through proposed mediators for belief in the spread of happiness and motive to strengthen relationships ...... 67

Table 10: Indirect effects and bias-corrected confidence intervals of IV on DV through proposed mediators for belief in the spread of sadness and motive to receive support ...... 68

Table 11. Hierarchical regression predicting nonverbal anger expressivity ...... 105

Table 12. Hierarchical regression predicting expected support ...... 106

Chapter 1: Introduction

Anne Frank said, “A person who is happy will make others happy.” She believed in the power of one person’s feelings to influence another’s. Abraham Lincoln, on the other hand, said,

“Most folks are about as happy as they make up their minds to be.” In his view, each person’s emotions were his own. We can look to science for evidence in support of these contrasting beliefs to determine which of the two was correct, but the accuracy of these statements is only one piece of the story. Merely holding one belief or the other may play an important role in guiding emotional behavior. Anne, believing it is in her power to make others happy, might be a bit quicker than Abraham to smile and to seek out others when she is feeling good.

Lay beliefs about emotions are important to study; they have been linked to emotional behavior, psychological health and well-being (De Castella et al., 2014; Mauss, Tamir,

Anderson, & Savino, 2011; Spokas, Luterek, & Heimberg, 2009; Tamir, John, Srivastava, &

Gross, 2007). Much of the literature on beliefs about emotions has emphasized beliefs at the intrapersonal level, such as whether one is able to control emotions (Tamir et al., 2007;

Williams, Chambless, & Ahrens, 1997) and whether emotions are valuable (Chow, Berenbaum,

& Flores, 2013; Mauss et al., 2011) or dangerous (Eifert, Zvolensky, Sorrell, Hopko, & W, 1999;

Eke & McNally, 1996; Ioannou & Fox, 2009; McCubbin & Sampson, 2006; Schmidt, Mitchell,

& Anthony, 2008; Schmidt, Zvolensky, & Maner, 2006; Smith, Wetterneck, Hart, Short, &

Björgvinsson, 2012). However, emotions are innately social—so much so that evolutionary theorists have proposed that they evolved in part to facilitate living in groups (Tooby &

Cosmides, 1990; Tooby & Cosmides, 1992). We experience our emotions with other people and because of them, and we often share our emotions with others (Rimé, Mesquita, Boca, &

Philippot, 1991). Emotions signal our feelings and interpersonal intentions (Knutson, 1996).

They help us to build and maintain our relationships (Fredrickson, 1998, 2004; Gable & Reis,

2010; Graham, Huang, Clark, & Helgeson, 2008). Emotions have real social consequences, but what do we know about how people conceptualize the interpersonal nature of emotions?

In this chapter, I will review the existing research related to beliefs about the interpersonal consequences of emotions. I will begin by summarizing research demonstrating that beliefs about emotions (i.e., emotion values) are partially grounded in emotions’ usefulness in pursuing social goals. Next, I will turn to the cultural psychological literature—tying cultural variation in beliefs about the value of emotions to culturally shaped social goals. Finally, I will discuss two areas of research in which beliefs about the social consequences of emotions have been examined in greater depth: happiness and social anxiety. In particular, these areas of research highlight that as people think about the social consequences of emotions, they may be primarily concerned with either themselves or others. In my dissertation research, I have attempted to contribute to this underdeveloped area of research by examining a novel construct: beliefs about the interpersonal spread of one’s emotions. Throughout this dissertation, I will explore the nature of these beliefs and their consequences for perceived and observable emotional behavior.

Beliefs about Emotions and Social Goals

As described above, emotions serve a host of social functions. As people pursue their social goals, perceptions of emotions’ social functions influence how they manage emotional experiences (Parrott, 1993; Tamir & Ford, 2012; Tamir, Ford, & Gilliam, 2013). People try to increase their anger during negotiations if they believe anger will be useful, especially when they

2 are motivated to perform well (Tamir & Ford, 2012b; Tamir et al., 2013). When participants in one study were made to believe that happiness is very valuable, they were more likely to try to increase their happiness during an emotion-eliciting film task, and to in turn feel more positive emotions during the films (Mauss et al., 2011). Similar patterns have been found in the case of sadness. Porat, Halperin, Mannheim, & Tamir (2016) found that people in Israel expressed a greater desire to feel sadness on Israeli National Memorial Day, a day commemorating the deaths of Israeli soldiers, if they believed it would help them fulfill their need to belong as a member of their group. This was true whether need to belong was measured as it naturally occurred or was manipulated to be more salient among some participants. People do not even need to be aware of their social goals for them to guide emotion preferences. Participants implicitly primed with a collaboration goal preferred activities that would make them feel happy and not activities that would make them feel angry, relative to those in a control group (Tamir,

Ford, & Ryan, 2013). Furthermore, experimental paradigms have demonstrated a causal role of beliefs in predicting emotion regulation—participants in another study who were led to believe anger was useful in negotiations preferred anger-eliciting activities and experienced anger more intensely than participants in a control condition (Tamir, Bigman, Rhodes, Salerno, & Schreier,

2015).

This research demonstrates that how people manage their emotions in social contexts depends upon their perceptions of emotions’ usefulness—people pursue emotions that are believed to be useful in attaining social goals. Beliefs about the overall value of emotions may be explained by the interpersonal goals one consistently pursues. I turn to the cultural psychological literature to examine this possibility. Cultural contexts have been shown to foster differing

3 emotion values. For example, while individuals across cultural contexts agree in their belief that positive emotions are preferable to negative emotions, the degree to which positivity is valued and negativity devalued varies across groups (Diener, Suh, Smith, & Shao, 1995). Social goals also differ across cultural contexts (e.g., Markus & Kitayama, 1991). Cultural psychologists studying beliefs about emotions have attempted to understand the relationships between these constructs. In the following section, I will begin by introducing some key areas of cultural differences in beliefs about emotions. Next, I will discuss the role of culturally shaped interpersonal goals in explaining variation in beliefs about emotions.

Culture and Beliefs about Emotions

Cultural differences in beliefs about emotions are well-documented. In the U.S., the dominant belief is that positive emotions can and should be maximized and negative emotions minimized (Diener et al., 1995; Eid & Diener, 2001; Ford, Shallcross, Mauss, Floerke, & Gruber,

2014; Koopmann-Holm & Tsai, 2014). Examples of this belief are pervasive: the constitution cites the pursuit of happiness as an inalienable right; the best-selling song of 2014, encouraged listeners to “clap along if [they] feel like happiness is the truth” (Billboard, 2014). Even sympathy cards encourage mourners to “hold on to hope” (Koopman-Holm & Tsai, 2014). In other cultures, however, individuals do not seem to value happiness and devalue unhappiness to the same extent as Americans. Instead, in many cultures the belief that happiness can sometimes be undesirable and unhappiness desirable is more prevalent (Chentsova-Dutton, Senft, & Ryder,

2013; Joshanloo & Weijers, 2014; Lu & Gilmour, 2004; Uchida & Kitayama, 2009; Uchida,

Norasakkunkit, & Kitayama, 2004).

Interestingly, recent research suggests that cultural differences in this domain are even more specific, fostering preferences for particular types of happy and unhappy experiences. For example, Jeanne Tsai and colleagues describe cultural differences in the particular positive emotions that are considered ideal. They find that people in Western cultural contexts tend to value high arousal positive emotions (i.e., excitement) more and low-arousal positive emotions

(i.e., calmness) less than do people in East Asian contexts, relative to other emotions (Tsai,

Knutson, & Fung, 2006; Tsai, Louie, Chen, & Uchida, 2007; Tsai, Miao, Seppala, Fung, &

Yeung, 2007). Other work finds that interpersonally disengaging emotions, like pride and anger, tend to be more highly valued in cultural contexts like the U.S., while interpersonally engaging emotions, like affection, shame, and guilt, are relatively more highly valued in cultural contexts like China and Japan (Eid & Diener, 2001; Kitayama, Markus, & Kurokawa, 2000; Kitayama,

Mesquita, & Karasawa, 2006b).

Beliefs about emotions have far-reaching consequences for emotion experience and expression, guiding individuals’ choices of activities and interactions (Sims, Tsai, Koopmann-

Holm, & Goldstein, 2014; Tamir, Mitchell, & Gross, 2008) and evoking particular types of emotional experiences (Boiger, Güngör, Karasawa, & Mesquita, 2014; Boiger, Mesquita,

Uchida, & Barrett, 2013; Kitayama et al., 2006b). Cultural differences in beliefs about emotions are evident quite early in life, as parents strive to instill culturally appropriate values and behaviors in their children (Cole, Tamang, & Shrestha, 2006; Fung, 1999; Keller et al., 2006;

Keller & Otto, 2009; Wang, 2001; Wang & Fivush, 2005). These beliefs also become represented outside of the head in the physical environment. They make their way into cultural products like religious texts, children’s books, even language, that are created by and

5 subsequently consumed by members of a cultural context (Dzokoto, Senft, Kpobi, &

Washington-Nortey, 2016; Tsai, Louie, Chen, & Uchida, 2007; Tsai, Miao, & Seppala, 2007).

As culturally shaped beliefs about emotions become represented in cultural products and parents’ socialization practices, these beliefs can be transmitted to new members and maintained over time.

Social Goals as Explanations of Cultural Variation

Emotions are critical to how we navigate our everyday social interactions and relationships, helping us to engage effectively with others. However, effective social engagement can look quite different across cultures. Emotions unfold within the context of culturally shaped ideas and practices that inform what it means to be a person and to engage with others (Boiger &

Mesquita, 2012). These cultural models influence how people think about their emotions as well as the relationships between beliefs about emotions and emotional experience, regulation, and behavior. Differing models of the self vis-à-vis relationships are perhaps the most commonly applied explanation of cultural differences in emotions. Markus & Kitayama (1991) proposed that in Western cultural contexts, an independent view of the self as distinct from others prevails.

Autonomy and uniqueness are highly valued. A person is best defined by the stable, internal, and unique traits that describe them. On the other hand, in East Asian cultural contexts an interdependent view of the self as inextricably tied to close others prevails. Interpersonal tasks such as adapting to situational demands and adjusting to accommodate the group are highly valued. Social relationships and group memberships are important components of individuals’ definitions of themselves.

These differing models of the self, and the social goals they engender, are often invoked to explain cultural variation in the emotion values described above. In independent cultural contexts like the U.S., emotions like pride, anger, and excitement are more highly valued than in other contexts because they align with the social goals of an independent self: to assert one’s individuality, to express oneself openly, and to influence others (Eid & Diener, 2001; Kitayama et al., 2000, 2006b). Pride is useful in differentiating the self from others. Anger is viewed as a mature, healthy means of expressing oneself (Shweder, Haidt, Horton, & Joseph, 2008), and excitement facilitates influencing others (Tsai, Miao, Seppala, Fung, & Yeung, 2007). The open expression of emotions is highly valued as a means of asserting independence and uniqueness

(Markus & Kitayama, 1991; Wierzbicka, 1992). In interdependent contexts, such as those in East

Asia, emotions like calmness, affection, and shame align with the social goals of an interdependent self: to attend to complex social cues and adjust to others’ expectations in pursuit of social connection and harmony (Eid & Diener, 2001; Kitayama et al., 2006b). Shame, for example, is considered useful in affirming one’s place in the social group by encouraging critical self-reflection and perspective taking (Heine, Lehman, Markus, & Kitayama, 1999). Calm states encourage attention to one’s environment and facilitate the adjustment to social norms and expectations that is values in interdependent contexts (Tsai et al., 2007). Emotional control is more highly valued than it is in interdependent contexts (Matsumoto, Yoo, & Fontaine, 2008;

Mauss, Butler, Roberts, & Chu, 2010).

In the research described above, beliefs about emotions are shaped by their actual social consequences: emotions are valued because they facilitate the achievement of culturally appropriate social goals. However, in studying the general value or utility of emotions, we are

7 limited in what we can infer about how individuals conceptualize the consequences of emotions.

Take the case of anger: the literature described above demonstrates that anger is valued more highly in the U.S. than in East Asian contexts because it is useful in asserting one’s independence. Further, people are more likely to try to feel angry when they perceive anger as useful. This is an important contribution to our understanding of emotions, but questions remain about the complex ways people may actually think through what anger means in social context.

In thinking that anger is valuable in a negotiation, are people primarily considering what anger means for themselves, for their interaction partners, or both? To gain insight into these questions, it is important that research explicitly assess beliefs about the social consequences of emotions.

Beliefs about the Social Consequences of Emotions

Beliefs about the social consequences of emotions are potentially quite complex, reflecting inherent complexity of social interactions. People may hold beliefs about how emotions will influence themselves (i.e., how they might influence others’ impressions of them).

People may also hold beliefs about how their emotions will impact others (i.e., how they will make others feel). Though these beliefs are understudied, some existing literature suggests that in independent cultural contexts like the U.S., where focus on the self is relatively more prominent, individuals’ beliefs about the social consequences of emotions tend to emphasize the consequences of emotions for themselves. In interdependent cultural contexts like those in East

Asia, salient concerns may be more likely to emphasize the consequences of one’s emotions for others. In this section, I will review two areas of literature that highlight these differing sets of concerns—work on happiness and social anxiety.

Happiness

As stated above, people in independent contexts like the U.S. tend to value happiness and devalue unhappiness more strongly than those in more interdependent cultural contexts. In East

Asian contexts, where social harmony is a primary goal, happiness tends to be defined in socially engaged ways—happiness is about feeling close to others and being relationally connected

(Ford, Dmitrieva, et al., 2015; Uchida & Kitayama, 2009; Uchida et al., 2004). Kitayama et al.

(2006) found that experiences of engaging emotions predict general positive feelings (i.e., happiness) among Japanese participants, while the experience of disengaging emotions like pride predicted general happiness among Americans. In pursuit of social harmony, East Asians are more likely to seek balanced, rather than uniformly positive affect (Uchida et al., 2004;

Schimmack, Oishi, & Diener, 2002). Lay theories of happiness in these contexts include its potential to yield negative social consequences, such as causing envy in others or making one inattentive to the demands of social situations. These concerns are rarely expressed as part of

American participants’ conceptions of happiness (Lu & Gilmour, 2004; Uchida & Kitayama,

2009; Uchida et al., 2004).

To infer from this research that people in individualistic cultures do not concern themselves with others is, of course, unjustified. Indeed, there are examples of social concerns being fairly prominent in this context. Miyamoto & Ma (2011) asked European American and

Japanese students to think about a time when they had dampened their positive emotions and to describe their reasons for doing so. Across cultures, the most common reason for trying to feel less positive emotions were social concerns—reported by 70% of Americans and 60% of

Japanese participants. However, this category included concerns about hurting others’ feelings as

9 well as about hurting one’s own social image. It may be that the primary motive for Americans was still fundamentally individualistic—to protect one’s social image. Additional data lend support to this hypothesis: Bastian and colleagues (2012) found that people in Australia, a relatively independent cultural context, were more likely than participants in Japan to believe others expected them to be happy, and that feeling unhappy would lead others to dislike them.

While people in interdependent cultural contexts may be primarily concerned with the consequences of happiness for others, people in independent cultural contexts may be relatively more concerned about social consequences for themselves.

These differing concerns have important implications for how people experience emotions. In line with their concerns about the costs of happiness for others, East Asians are more likely to experience mixed emotions (Grossmann, Huynh, & Ellsworth, 2015; Leu et al.,

2010; Schimmack et al., 2002; Spencer-Rodgers, Peng, & Wang, 2010), especially following positive experiences (Miyamoto & Ma, 2011; Miyamoto, Uchida, & Ellsworth, 2010). Miyamoto et al. (2010) demonstrated that concerns about the social consequences of happiness, including its potential to cause unhappiness in others, led East Asians to view positive situations as bittersweet, while European Americans viewed the same situations as uniformly positive. For

Americans, concerned that lack of happiness may spur negative social evaluation, there may be little motivation to see the downsides of positive events.

These beliefs also have implications for long-term mental health and well-being. In fact, these differing social concerns may help explain a paradox in the happiness literature: that valuing happiness is sometimes associated with improved well-being (e.g., Bastian, Kuppens, De

Roover, & Diener, 2014) but sometimes with decreased well-being (Ford, Mauss, & Gruber,

2015; Ford et al., 2014; Mauss et al., 2012, 2011). Ford, Dmitrieva, et al., (2015) demonstrated that the motivation to pursue happiness was associated with decreased well-being among participants in the U.S., but increased well-being among participants in interdependent Russian and East Asian contexts, and the tendency for those in interdependent contexts to define happiness in terms of social engagement explained these differing relationships. The authors suggest that conceptualizing happiness in terms of social engagement may mean that those who are motivated to pursue happiness do so in more socially engaged ways—investing time in their relationships and behaving prosocially towards others. For Americans, on the other hand, the concern about protecting one’s social image by maintaining happiness may exacerbate the experience of negative emotions. People believing others expected them not to feel negative emotions actually experienced more frequent and intense feelings of sadness (Bastian et al.,

2012). In a follow-up study, Bastian et al. (2015) found that those who experienced frequent negative emotions and believed those emotions were socially unacceptable felt lonelier. The authors suggest that for people who believe others expect them to be happy, negative emotions are compounded by feelings of failure for not meeting others’ expectations and of concern that others will think poorly of them.

This research demonstrates that social concerns about happiness are prevalent among members of independent and interdependent cultures. However, the primary referents of these concerns seem to differ across contexts. People engaged in interdependent contexts are more likely to believe that their happiness may have negative consequences for the people around them. Given that a primary social goal in these contexts is relational harmony, they are more likely to regulate their positive emotions accordingly. Meanwhile, people engaged in

11 independent contexts are more likely to believe that others will evaluate them negatively if they are not happy. In this context, people are motivated to protect their social image by maintaining positivity—which can exacerbate the experience of negative emotions when they inevitably occur.

Social Anxiety

The literature on social anxiety provides an additional context for contrasting concerns about the social consequences of emotions for oneself vs. others. Literature on Social Anxiety

Disorder (SAD) in the U.S. suggests that concerns about the consequences of emotions for oneself are central to the disorder, consistent with an independent view of the self. Indeed, the disorder is characterized by an intense concern over public scrutiny and fear of social rejection

(American Psychiatric Association, 2013). Research conducted primarily in the U.S. has found that people with social anxiety disorder are more likely to believe that emotions are a sign of weakness and that they lead to negative evaluation and social rejection (McWilliams, Stewart, &

MacPherson, 2000; Spokas et al., 2009). In contrast, work in the area of cultural-clinical psychology has identified cultural variations of SAD in East Asian cultural contexts that are characterized by concern for others, consistent with an interdependent view of the self. Taijin

Kyofusho (TKS), like SAD, includes symptoms of distress and anxiety surrounding social situations. Unlike SAD, however, the primary concern of those with TKS is that they will offend or embarrass others (Kleinknecht, Dinnel, Kleinknecht, Hiruma, & Harada, 1997). People with this symptom may worry, for example, that they will embarrass others if they blush, or irritate others if they display nervousness (Sasaki & Tanno, 2006). Sasaki, Wada, & Tanno (2013) found concerns about offending others were also central to the experience of Egorrhea, an anxiety

12 disorder first described in Japanese clinical settings. Egorrhea is characterized by the belief that one’s internal states, including feelings, leak out uncontrollably and can be seen by others. The authors found this belief was not uncommon or particularly distressing in normal Japanese populations. However, believing one’s internal states could not be controlled was distressing and sometimes pathological when people believed their internal states were offensive to others

(Sasaki et al., 2013, study 3).

Interestingly, some research shows that symptoms of SAD, TKS, and Egorrhea may overlap substantially, with one study finding 75% of patients diagnosed with SAD in the U.S. endorsed at least one symptom typically associated with TKS (Choy, Schneier, Heimberg, Oh, &

Liebowitz, 2008; Hofmann, Asnaani, & Hinton, 2010; Sasaki et al., 2013). There are many reasons why these other-focused symptoms may be common but not reflected in rates of diagnosis in independent cultural contexts like the U.S. Clinicians’ culturally shaped beliefs about the nature of anxiety or knowledge of cultural variants of the disorder may bias the symptoms they assess; alternately, patients may spontaneously focus on the symptoms they view to be of primary concern (Tseng, Asaim, Kitanishi, Mclaughlin, & Kyomen, 1992; Choy et al.,

2010).

As was the case with social concerns surrounding happiness, the concerns associated with social anxiety have important implications for emotional experience and behavior. Research conducted in the U.S. finds that people with SAD are more likely to try and hide their emotions from others (Campbell-Sills, Barlow, Brown, & Hoffman, 2006; Spokas et al., 2009). Spokas et al. (2009) found that the belief that emotions were a sign of weakness partially mediated the relationship between anxiety disorder and reported use of expressive suppression. In a laboratory

13 setting, Campbell-Sills et al. (2006) found that individuals with anxiety disorders were more likely to suppress their emotions than non-clinical participants following a negative emotion- eliciting film clip. This relationship was mediated by the participants’ belief that their negative emotions were unacceptable. These relationships are particularly troubling in independent cultural contexts, where expressive suppression has been consistently associated with poorer short and long term emotional, social, and cognitive outcomes (Butler, Lee, & Gross, 2007;

Gross, 2002; Haga, Kraft, & Corby, 2009; Mauss, Bunge, & Gross, 2009).

As was the case with beliefs about the social consequences of happiness, the social anxiety literature suggests that social concerns may take different forms depending upon one’s cultural context. In independent cultural contexts, the experience of social anxiety is primarily characterized by self-presentation concerns. In the case of East Asian social anxiety, it seems that relevant concerns are more likely to center around others, with patients worrying that they might offend or embarrass others.

Summary and Limitations

The studies reviewed in this chapter suggest that beliefs about which emotions are valuable emerge from culturally salient interaction goals. Socially disengaging and high-arousal positive emotions, which are useful in influencing others and facilitating the assertion of individual selves, are perceived to be appropriate and valuable in Western, independent cultural contexts. Socially engaging and low-arousal positive emotions, which facilitate adjustment to social situations, are perceived as more appropriate in interdependent East Asian context. These beliefs about emotions provide insight into variation across individuals and cultures in actual emotional experience and behavior, as well as longer-term well-being outcomes. However, this

14 literature is also limited. The social consequences of emotions are complex; they include consequences for oneself as well as others. Research in the literatures on happiness and social anxiety suggests that culture may shape which of these individuals focus on when they conceptualize the social consequences of their emotions. It seems that for people engaged in independent cultural contexts, even social concerns may be fundamentally self-focused. In these contexts, people are more likely to cite concerns related to social evaluation or their own feelings of embarrassment. For people in cultural contexts where an interdependent view of the self prevails, social concerns seem more likely to actually focus on consequences of emotions for others.

The existing research demonstrates the importance of including beliefs about emotions in models of emotion, and of taking cultural factors into account when examining such beliefs and their relationships to emotional experience, behavior, and well-being. However, the research explicitly examining beliefs about the social consequences of emotions is underdeveloped, despite the highly interpersonal nature of emotions. As a result, we know little about how people conceptualize the consequences of their emotions for others, and how these beliefs influence emotional experience and behavior.

Current Research

In the chapters to follow, I will describe three studies in which I examine a novel construct: beliefs about the spread of emotions to others. The idea that emotions spread—that one’s happiness can make people feel good and one’s sadness can bring them down—appears fairly often in cultural products and lay dialogues. Coca Cola had an entire ad campaign titled

“spread happiness.” A popular Calvin and Hobbes cartoon closes with the insight that, “Nothing

15 helps a bad mood like spreading it around.” In everyday conversations, I have heard friends and family give voice to a range of beliefs about what their emotions mean for others. Some are so convinced that everyone is in agreement that emotions do spread that they are confused about why I study the question at all. Others seem less convinced—I had a phone call with a family member recently who recounted how she had been feeling upset all day by a string of setbacks, then added that to top it off, her boyfriend was “in such a bad mood all of sudden.”

In truth, emotions do seem to transmit in social contexts. People automatically mimic others’ emotional behaviors and, in turn, “catch” others’ emotional experiences (Hatfield,

Cacioppo, & Rapson, 1994). Over time, the emotional experiences of people that spend time together become more similar (Anderson, Keltner, & John, 2003). The term emotional contagion refers to this phenomenon of catching others’ emotions, and past research has shown that self- reports of one’s susceptibility to catching others’ emotions are associated with outcomes like emotionality, and predict emotional responses to others’ emotional expressions (Doherty, 1997).

Though related, one’s susceptibility to catching others’ emotions is conceptually distinct from beliefs about whether one’s emotions will spread to others. Doherty states, “ Susceptibility is best considered the tendency to automatically mimic and synchronize with the expressions of others and, through afferent feedback from the facial and/or skeletal muscular activity, to experience or

“catch” the others' emotions (Doherty, 1997, p. 149).” Endorsement of the belief that emotions spread is best considered an enduring assumption that one’s own emotions transmit to others in interpersonal context. No assertion is made as to the assumed mechanism behind this spread—an individual may believe their emotions spread through some automatic process (as described by

Doherty above), through one’s own agency, through others’ aggregated susceptibilities, or any

16 combination of these. Some individuals may be quite susceptible to others’ emotions, yet not believe that others share this tendency, or that they themselves have the agency to transmit their emotions. Though the weight of research suggests emotions do spread, beliefs about psychological phenomena are not always accurate, and even inaccurate beliefs can be influential.

As far as I am aware, only one existing study has examined this construct explicitly.

Manser, Cooper, & Trefusis (2012) developed a scale to measure beliefs about emotions that the authors proposed would be implicated in emotion dysregulation and clinical symptoms. The authors measured a range of potentially problematic beliefs—that emotions are uncontrollable, shameful, meaningless, useless, damaging, and contagious. The belief that one’s negative emotions were contagious was significantly correlated with measures of borderline symptoms, depression, and anxiety in a non-clinical sample. Most interestingly for the purposes of the current research, people that endorsed the idea that their negative emotions were contagious were also more likely to say they cried and expressed their feelings in order to cope with their strong feelings. This research demonstrates that there may be some associations between people’s beliefs about the spread of emotions, at least of negative emotions, and people’s perceptions of their own emotional expressivity. However, this research was clinically focused, and these items situated amid a range of measures clearly assessing negative beliefs about emotions. The belief that emotions spread, however, is not an explicitly negative one. The spread of emotions may be perceived to be either good or bad, depending upon how people perceive its consequences.

Beliefs about the Spread of Emotions and Social Goals

Beliefs about the spread of emotions may influence how people engage in interpersonal emotion regulation—efforts to manage their own or others’ affective experiences (for a review,

17 see Zaki & Williams, 2013). People may aim to influence others’ emotions for a number of reasons, but here I will focus on four goals ranging in the degree to which they primarily focus on the self: the goal to benefit others, to strengthen relationships, and to receive support.

It is important to acknowledge that while goals may be more explicitly focused on others or the self, the two are closely related in actual interactions. For example, as the literature on altruism highlights, one may be motivated to help others but, in doing so, also benefit oneself

(see Batson, 1991). Similarly, one may want to elicit support from a friend but also recognize that doing so impacts the friend as well. While recognizing this complexity, I sampled motivations that differ in the relative emphasis on others versus the self. While I also recognize there may be additional constructs that are relevant to understanding these relationships, I focus on these particular motives because past research has established their relevance to interpersonal emotion regulation and because they can be theoretically linked to belief in the spread of emotion. Thus, these constructs provide a rich starting point for the current investigations, which aim to identify and explain the links between belief in the spread of emotions and emotional behavior.

Other-Focused Goals: To Benefit Others and Strengthen Relationships

Beliefs about the spread of emotions may influence emotional behavior through their relationships to self- or other-focused social goals. Here, I will outline potential relationships between beliefs about the spread of emotions, social goals, and emotional behavior. First, people may aim to influence others’ emotions to benefit (or avoid harming) them (Batson, Duncan,

Ackerman, Buckley, & Birch, 1981; Goetz, Keltner, & Simon-Thomas, 2010). If the goal is to benefit others, people may try to increase others’ positive and decrease their negative emotions.

Believing that one’s emotions spread will likely be associated with increased positive expressivity (which will spread positive feelings) but decreased negative expressivity (which will spread negative feelings). Next, people may be motivated to strengthen their relationships with others, a goal that may serve both members of an interaction. Sharing in others’ emotional experiences promotes affiliation, helping to build and strengthen relationships (Collins & Miller,

1994; Glover & Parry, 2008; McPherson, Smith-Lovin, & Cook, 2001). Thus, if one’s goal is to strengthen relationships, belief that emotions spread is likely to be associated with increases in both positive and negative emotional behavior.

Self-Focused Goals: To Receive Support

Alternately, people may have more self-focused motivations for altering others’ emotional experiences. As Netzer, Van Kleef, & Tamir (2015) demonstrate, people will attempt to make others feel badly if doing so will be useful in achieving their goals. People may view the spread of emotions as a helpful means of getting others to respond supportively. After all, those that actively share in one’s happiness or negative emotions tend to be the most effective supportive providers (Dakof & Taylor, 1990; Gable & Reis, 2010; Han et al., 2011; Thoits, 2011; though see also Nils & Rimé, 2012). Thus, if one’s goal is to receive support from others, belief in the spread of emotions will likely be associated with increases in both positive and negative emotional behavior. In summary, the relationship between belief in the spread of emotions and emotional behavior may differ depending upon one’s social goals.

Social Goals in Cultural Context

Finally, culture plays an important role in shaping social goals. South Korea is a fitting cultural context to contrast to the U.S. in examining these constructs. Like other East Asians,

Koreans are relatively less likely than Americans to endorse an independent view of the self.

South Koreans also tend to report greater concern over the potential consequences of their emotions for others than European Americans. In particular, Koreans report being more concerned about burdening others with their negative feelings, and these concerns explain a reduced tendency among Koreans to seek social support from others (Kim, Sherman, Ko, &

Taylor, 2006; Taylor et al., 2004).

Relationships between beliefs about the spread of emotions and expressive behavior may be similar across groups. If this is the case, examining these relationships in contexts with differing social goals allows for a more rigorous test of the generalizability of any such relationships. Relationships between beliefs about the spread of emotions and expressive behavior may also differ across cultures. In particular, these relationships may depend more heavily upon other-focused goals among Koreans, and self-focused goals among Americans.

Examining these relationships in contexts previously shown to differ in their tendencies to endorse such goals maximizes the ability to detect cultural differences if they do exist.

Summary of Studies

Across the three studies presented here, I aim to explore the nature of beliefs about the spread of emotions and their relationships to emotionally expressive behavior. In Study 1, I validate a new scale to measure beliefs about the spread of one’s happiness and sadness. I describe the nature of these beliefs and explore their relationships to participants’ perceptions of their own emotional behavior. In Study 2, I seek to identify the mechanisms underlying relationships between beliefs about the spread of emotions and people’s perceptions of their own emotional behavior. I also examine whether the structure of beliefs and their relationships to

20 reported emotional behavior are similar or different in the U.S. and South Korea. Finally, Study

3 uses an experimental approach to examine the causal nature of relationships between beliefs about the spread of emotions, social goals, and emotional behavior in the U.S. and South Korea.

Chapter 2: Understanding the Nature of Beliefs about the Spread of Emotions and

Exploring their Relationships to Reported Emotional Behavior

In the current chapter, I examine the nature of beliefs about the spread of emotions and the relationships between these beliefs and reported emotional behaviors. In Study 1A, I created the Beliefs About the Spread of Emotions (BASE) scale to measure beliefs about the spread of happiness and of sadness and conducted an exploratory factor analysis to determine the structure of these beliefs. Study 1B worked to clarify the nature of these constructs. Towards this end, I first distinguish beliefs about the spread of emotions from potentially related constructs, such as the tendency to catch emotions from others (i.e., emotional contagion), or affective traits related to temperament (i.e., Big Five, trait affect). Next, I examine associations between beliefs about one’s own ability to spread happiness and sadness to others (as measured by the BASE scale) with an additional measure assessing more generalized beliefs about the spread of a wider range of emotions to particular groups of people. These tests are meant to clarify whether responses to the BASE scale can be generalized to other positive and negative emotions and whether they are specific to particular groups of people. Finally, I explore relationships between beliefs about the spread of one’s emotions and participants’ reports of their own verbal and nonverbal emotional behavior.

Study 1A

Method

Participants. Students ( n = 114) at a Jesuit university on the East Coast participated in this study for course credit. Approximately 70% of the sample was female ( n = 80). Participants

22 were, on average, 20.16 years old ( SD = 3.90). Information on race/ethnicity was not collected in this sample of participants.

Measure: Beliefs about the spread of emotions (BASE). I developed a 20-item measure composed of 10 items each assessing the degree to which participants believed their happiness and sadness spread to others (e.g., “My happiness brightens others’ moods”, “My sadness rubs off on other people”). To develop items, I took note of the language people use when talking about the idea of emotional contagion in every day conversations. I also read a number of blogs and popular-press articles in which people discussed ideas relating to the spread of emotions, taking note of the language that writers used in communicating about these topics.

For example, an article in Psychology Today states, “when you hang out with happy people, you tend to feel happier,” and “One family members' depression can bring down an entire family

(Bourg Carter, 2012).” In a blog called Tiny Buddha, the author writes, “Smile. This is the best method ever to…spread happiness around to others (Kristi Frimpong, 2013).” Items were rated on a 7-point scale (1 = Strongly disagree, 7 = Strongly agree, see Table 1).

Procedure. Participants completed a series of online surveys, including the BASE scale, in exchange for course credit. Participants also completed a range of measures related to cultural values that are unrelated to the present study. Online surveys were completed outside of the laboratory, in whichever location was most convenient for participants.

Results

Factor analysis. Because some of the items in the scale were non-normally distributed, the data were analyzed using the principal axis factoring extraction method; direct oblimin rotation was used to account for correlation among the factors (Fabrigar, Wegener, MacCallum,

& Strahan, 1999). Evaluation of the scree plot suggested a two-factor solution (see Table 1). The first factor, comprised of the 10 items assessing beliefs about the spread of happiness, explained

37.54% of the variance. Factor loadings for the first factor ranged from .51 to .90. The second factor, corresponding to the 10 items measuring beliefs about the spread of sadness, explained an additional 22.58% of variance. Factor loadings ranged from .64 to .83. Internal reliability was high for both factors ( α = .94 for happiness, α = .93 for sadness). Beliefs about the spread of happiness and sadness were positively correlated with each other ( r (112) = .25, p = .007). On average, individuals tended to endorse the idea that their happiness spread ( M = 5.14; SD = .87) to a greater degree than they did the idea that their sadness spread ( M = 4.40, SD = .98; t (113) =

6.98, p < .001).

Discussion

Factor analysis produced two factors distinguishing beliefs about the spread of happiness and beliefs about the spread of sadness, suggesting that lay theories about the spread of emotions are at least partially emotion-specific. Zero-order correlations further indicate that these beliefs are positively correlated with each other, yet are separate constructs. Finally, lay beliefs in the spread of happiness seem to be endorsed more strongly than beliefs in the spread of sadness.

Study 1B

Method

Participants. Students at a Jesuit university on the East Coast (n = 129) participated in this study for course credit, and Mechanical Turk workers ( n = 209) participated in this study for payment. Information on the gender, age, and race of participants in each of these samples can be found in Table 2. The racial compositions of these samples did not significantly differ ( X2 (4, N =

337) = 4.02, p = .40). However, mTurk workers were, on average, older ( t (336) = 13.23, p <

.001) and marginally more likely to be male ( X2 (1, N = 337) = 3.71, p = .061). Analyses confirmed that relationships between variables did not differ across these two samples when age and gender were controlled, therefore the samples were combined.

Procedure. Participants completed an online survey in whichever location they found most convenient. Participants first completed the BASE scales, as described in Study 1A.

Participants also completed a range of measures meant to assess their perceptions of their own emotional behavior as well as measures of their trait affect, Big Five personality dimensions, tendency to catch others’ emotions, and generalized beliefs about the spread of emotions.

Participants reported their demographic characteristics at the end of the survey.

Measures.

Beliefs about the spread of emotions (BASE). Participants completed the 20-item BASE scale as described in Study 1A. The 20-item measure includes two 10-item subscales assessing the degree to which participants believe their happiness spreads to others ( α = .90) and their sadness spreads to others ( α = .91). Items were rated on a 7-point scale (1 = Strongly disagree, 7

= Strongly agree).

Emotional contagion . The 15-item Emotional Contagion Scale (Doherty, 1997) assesses propensity to catch others’ emotions (e.g., “If someone I'm talking with begins to cry, I get teary- eyed,” α = .84). Items were rated on a 5-point scale (1 = Never, 5 = Always).

Trait affect . The Positive and Negative Affect Schedule (PANAS, Watson, Clark, &

Tellegen, 1988) measured the extent to which participants felt ten positive emotions (e.g.,

“interested,” “enthusiastic;” α = .90) and ten negative emotions (e.g., “distressed,” “nervous;” α

= .91) over the past week. Items were rated on a 5-point scale (1 = Very slightly or not at all, 7 =

Extremely).

Big-five personality dimensions . The student sample, but not the mTurk sample, completed a measure of personality. The Ten-Item Personality Inventory (TIPI, Gosling,

Rentfrow, & Swann Jr., 2003) assessed the degree to which participants believed pairs of traits corresponding to extraversion ( α = .78), agreeableness ( α = .45 ), conscientiousness ( α = .69), neuroticism ( α = .56), and openness ( α = .51) applied to them. Items were rated on a 7-point scale (1 = Strongly disagree, 7 = Strongly agree). Because of the low internal reliability of the agreeableness, neuroticism, and openness subscales, the associations with these constructs that are reported below should be interpreted cautiously.

Generalized beliefs about the spread of emotions. An additional measure was developed to assess the degree to which participants believed three negative emotions (anxiety, sadness, and anger) and three positive emotions (calmness, excitement, happiness) spread to ingroup members

(i.e., family, close friends) and outgroup members (i.e., acquaintances, strangers). Participants were asked, for example, “When a person feels anxious, how likely do you think it is that their anxiety will spread to others when they are with each of the following groups of people?” Items were rated on a 5-point scale (1 = Very slightly or not at all, 5 = Extremely). Scales were formed assessing beliefs about the spread of positive emotions to ingroup members ( α = .87) and outgroup members ( α = .87) and about the spread of negative emotions to ingroup ( α = .87) and outgroup members ( α = .86).

Emotion regulation. The Emotion Regulation Questionnaire (Gross & John, 2003) assessed participants’ use of emotion regulation strategies. Of interest to the current research, a

4-item subscale measures use of expressive suppression (e.g., “I control my emotions by not expressing them;” α = .75). A 6-item subscale assesses use of cognitive reappraisal (e.g., I control my emotions by changing the way I think about the situation I'm in”). However, because this regulatory strategy is less directly associated with emotional behavior, it is not discussed in the current research. Items were rated on a 7-point scale (1 = Strongly disagree, 7 = Strongly agree).

Nonverbal emotional behavior. The Berkeley Expressivity Questionnaire (BEQ, Gross

& John, 1995) assessed participants’ perceptions of their own nonverbal emotional expressivity.

This scale includes a 4-item subscale assessing positive expressivity (e.g., “When I'm happy, my feelings show;” α = .73) and a 6-item subscale assessing negative expressivity (e.g., “No matter how nervous or upset I am, I tend to keep a calm exterior,” reverse coded; α = .75). Items were rated on a 7-point scale (1 = Strongly disagree, 7 = Strongly agree).

Distress disclosure . The 12-item Distress Disclosure Index (Kahn & Hessling, 2001) assessed the degree to which participants report talking to others about their distress (e.g., I try to find people to talk with about my problems;” α = .94). Items were rated on a 7-point scale (1 =

Strongly disagree, 7 = Strongly agree).

Capitalization. A 4-item measure of Capitalization (Lambert, Gwinn, Stillman, &

Fincham, 2011) assessed the degree to which participants capitalize on their positive experiences by sharing them with others (e.g., “I’m constantly telling people my good news;” α = .81). Items were rated on a 7-point scale (1 = Strongly disagree, 7 = Strongly agree).

Results

Discriminant validity. Correlations were calculated using Spearman’s rho to allow for potential non-linear relationships between the happiness and sadness subscales of the Beliefs

About the Spread of Emotions Scale (BASE) and emotional contagion (ECS), trait affect

(PANAS), and Big Five (TIPI; see Table 3). As expected, both subscales were significantly associated with, but distinct from, emotional contagion (BASE-Happy: r (338) = .33, p < .001;

BASE-Sad, r (338) = .31, p < .001). In the domain of trait affect, belief in the spread of happiness was related to, but not identical to, both trait positive affect ( r (338) = .46, p < .001) and negative affect ( r (338) = -.24, p < .001). Belief in the spread of sadness, however, was related to trait positive ( r (338) = .19, p < .001) but not negative affect ( r (338) = -.03, p = .575).

In regards to the Big Five, beliefs about the spread of happiness were positively related to extraversion ( r (129) = .27, p = .002), agreeableness ( r (129) = .23, p = .007), and openness ( r

(129) = .19, p = .028) and were negatively correlated to neuroticism ( r (129) = -.19, p = .028).

Beliefs about the spread of sadness were associated only with extraversion ( r (129) = .27 p =

.005). Neither belief was significantly associated with conscientiousness ( p’s > .10). In sum, beliefs in the spread of happiness and sadness are associated with emotional contagion, trait affect (especially positive affect), and the Big Five (especially extraversion). However, these correlations are in the weak to moderate range, validating that the BASE scales measure conceptually distinct constructs; this establishes discriminant validity.

Relationships to generalized beliefs. Correlations using Spearman’s Rho were calculated between the BASE scales and generalized beliefs about the spread of positive and negative emotions to ingroup and outgroup members. As seen in Table 3, beliefs in the spread of

28 one’s own happiness were most strongly correlated with generalized beliefs about the spread of positive emotions to ingroup members ( r (338) = .46, p < .001), but only weakly associated with beliefs about the spread of positive emotions to outgroup members ( r (338) = .14, p = .012).

Similarly, beliefs in the spread of one’s own sadness were most strongly correlated to generalized beliefs about the spread of negative emotions to ingroup members ( r (338) = .42, p <

.001), but only weakly related to beliefs about the spread of negative emotions to outgroup members ( r (338) = .12, p = .035).

Relationships to reported emotional regulation and behavior. I next assessed the relationships between beliefs about the spread of emotions and five reported emotional behaviors

(expressive suppression, positive nonverbal expressivity, negative nonverbal expressivity, capitalization, and distress disclosure) using hierarchical linear regressions. All control variables were entered in the first step. Trait positive and negative affect were included in order to ensure relationships observed between beliefs and emotional behaviors were not merely artifacts of the association between beliefs and typical emotional experience. Emotional contagion was included to ensure beliefs about the spread of emotions contributed predictive power beyond that provided by one’s own tendency to catch others’ emotions. Analyses also controlled for age and gender. In the second step, both the happiness and sadness subscales of the BASE scale were entered in order to assess their independent effects. The third step tested whether age or gender interacted with beliefs about the spread of happiness or sadness in predicting reported behavior. Finally,

Step 4 tested whether beliefs about the spread of happiness and sadness interacted with each other in predicting reported behavior. Neither Step 3 nor step 4 contributed significantly to the explanatory power of the models, indicating that relationships between beliefs and reported

29 behaviors were consistent across age and gender, and that beliefs about the spread of happiness and sadness do not significantly interact with each other. However, the inclusion of beliefs about the spread of emotions in Step 2 significantly contributed to the explanatory power of the models, indicating that these beliefs predicted reported behaviors beyond what could be accounted for through age, gender, trait affect, or emotional contagion.

Beliefs about the spread of happiness. As shown in Table 4, the belief that one’s happiness spreads predicted decreased reported use of expressive suppression ( β = -.17, t (329) =

-2.74, p = .01), increased reported positive nonverbal behavior ( β = .13, t (329) = 2.55, p = .01), and increased capitalization ( β = .13, t (329) = 2.37, p = .03), but did not significantly predict reported negative behavior ( β < .01, t (329) = 0.04, p = .97), or distress disclosure ( β = -.03, t

(329) = -.53, p = .60). Neither Step 3 nor step 4 contributed significantly to the explanatory power of the models, indicating that relationships between beliefs and reported behaviors were consistent across age and gender, and that beliefs about the spread of happiness and sadness do not significantly interact with each other.

Beliefs about the spread of sadness. Belief in the spread of sadness significantly predicted decreased reported use of expressive suppression ( β = -.11, t (329) = -1.99, p = .05) and increased reports of negative nonverbal expressivity ( β = .14, t (329) = 2.61, p = .01), capitalization ( β = .11, t (329) = 2.20, p = .03), and distress disclosure ( β = .17, t (329) =

3.18, p < .001), but not positive nonverbal expressivity (β = .01, t (329) = 0.18, p = .86),

Discussion

The aim of the current study was to gain insight into the nature of beliefs about the spread of emotions and the relationships between these beliefs and people’s perceptions of their own

30 emotional behavior. First, I tested the degree to which beliefs in the spread of happiness and sadness were associated with potentially related constructs. Beliefs in the spread of emotions are associated with, yet distinct from, self-reported tendencies to catch others’ emotions, trait affect, and the Big-5 personality constructs. For both beliefs, associations with emotional contagion, trait affect, and the Big-5 were low to moderate in magnitude, indicating that these are distinct constructs. In particular, people that believe their happiness or their sadness spread are also more likely to say that they tend to catch emotions of people around them. Both sets of beliefs are also associated with trait positive affect. It is possible that people who tend to experience positive emotions are more attentive to the impacts of their emotions on others, as positive emotion has been associated with broadened attention in past research (Fredrickson, 1998). Belief in the spread of happiness, but not sadness, was associated with decreased reported trait negative affect.

The lack of relationship between belief in the spread of sadness and trait negative affect may be due to the fact that the PANAS scale measures activated negative emotions, such as anxiety, but not low-activation negative emotions, such as sadness. People that believed their emotions spread also tended to be more extraverted. This is not surprising; extraverts may believe they can influence others more than introverts. Belief that one’s happiness spreads was further associated with agreeableness, openness, and decreased neuroticism. It may be that the spread of happiness is perceived as generally desirable, and as such is associated with other broadly desirable traits.

However, relationships (or lack thereof) with agreeableness, openness, and neuroticism should be interpreted cautiously, given the low internal reliability of these scales in the current sample.

Next, to gain insight into the generalizability of these beliefs across emotions and social contexts, I assessed relationships between the BASE scales and more generalized beliefs about

31 the spread of happiness, excitement, calmness, sadness, anxiety, and anger to ingroup members

(i.e., family and friends) and outgroup members (i.e., strangers and acquaintances). Beliefs about the spread of one’s own happiness most strongly correlated with generalized beliefs about the spread of positive emotions to ingroup members, while beliefs about the spread of one’s own sadness to others were most strongly correlated with generalized beliefs about the spread of negative emotions to ingroup members. I infer from these results that when people complete the

BASE scales, they are more likely to be thinking about the effects of their emotions on ingroup members, such as close friends or family, rather than on outgroup members, such as acquaintances or strangers. These data also provide some preliminary evidence that beliefs about the spread of happiness may generalize to beliefs about the spread of other positive emotions, and likewise beliefs about the spread of sadness may generalize to other negative emotions.

Finally, the current study aimed to explore whether beliefs about the spread of emotions are related to perceptions of one’s own emotional behavior. Beliefs in the spread of happiness and sadness each independently predicted reported emotional behaviors, beyond what was predicted by tendencies to experience positive and negative affect as well as participants’ own susceptibility to catching others’ emotions. More specifically, the more people believe their happiness spreads, the less they report using expressive suppression to regulate their emotions and the more they report expressing their positive emotions nonverbally as well as by talking about their positive experiences with others. The more people believe their sadness spreads, the less they report using expressive suppression, the more they report expressing their negative emotions nonverbally, and the more they report discussing their positive and negative emotional experiences with others. Results suggest beliefs about the spread of happiness may be most

32 closely associated with beliefs about one’s positive emotional behaviors. Beliefs about the spread of sadness are likewise more closely and consistently related to reported negative emotional behaviors, though a significant relationship between this belief and capitalization suggest that it may be may also be associated with some positive emotional behaviors. These patterns of relationships were quite robust across ages and genders—relationships between beliefs in the spread of emotions and reported behavior were not significantly influenced by respondents’ age or gender.

Limitations

Some of the results of Study 1 were surprising. While it makes intuitive sense that believing one’s happiness spreads might motivate more positive emotional behavior, I was surprised to find that those who believe their sadness spreads are more likely to report expressing their negative emotions. I proposed that belief in the spread of emotion might influence emotional behavior through its relationship with social goals. Namely, individuals may hold other-focused social goals to make others feel good or to strengthen relationships, or self-focused goals to receive high quality support or to assert their uniqueness. The relationships between belief in the spread of happiness and greater reported positive expressivity might be explained by motivations to benefit others, strengthen relationships, or receive support, as each of these goals is more likely to be achieved when happiness is shared. It is possible that belief in the spread of sadness is related to multiple motivations. For example, those that believe in the spread of sadness may be motivated to receive high-quality support or strengthen relationships, leading to increased reported expression of negative emotions, but also motivated to make others feel good, leading to increased reported expression of positive emotions. However, because these potential

33 explanations were unmeasured in Study 1, it is not possible to make assumptions about the underlying mechanisms behind these relationships.

I was also surprised by the consistency of the relationships reported above across age groups and genders, yet recognized the importance of testing the extent to which these results generalize to other groups. The relationships between beliefs about the spread of emotions and reported emotional behaviors may differ in cultural contexts where salient interpersonal goals differ from those of the United States. In the U.S., an independent view of the self as distinct from others tends to prevail (Markus & Kitayama, 1991). On the other hand, in East Asian cultural contexts like South Korea, an interdependent view of the self as tied to others tends to be more common (Markus & Kitayama, 1991). These differing models of the self have been used to explain cultural variation across a range of variables related to emotional experience and expression. For example, those with an interdependent view of the self are more likely to adjust to others’ expectations of them during emotional situations (Boiger, Mesquita, Tsai, & Markus,

2012) and to view emotions that foster connection among individuals as more desirable (Eid &

Diener, 2001; Kitayama, Mesquita, & Karasawa, 2006a). Koreans, in particular, are less likely to seek social support due to heightened salience of social concerns—the potential for one’s emotions to be burdensome to others or to damage relationships (Taylor et al., 2004). If Koreans are more concerned than Americans with the potential negative consequences of their emotions for others, Korean participants should report expressing their emotions less when they believe they will spread. However, if Koreans are more motivated by the receipt of social support or strengthening relationships, the relationships between beliefs in the spread of emotions and emotional behavior to be similar to those found among American participants in this study.

Table 1. Exploratory factor analysis of the BASE scale Item Happiness Sadness h2 When I am happy, others are unlikely to change their mood (rc) 0.51 0.29 0.29 When I am happy, people around me become happier 0.88 0.22 0.77 My happiness brightens others' moods 0.89 0.18 0.79 My happiness isn't infectious (rc) 0.76 0.19 0.58 The people around me become happier when I'm happy 0.90 0.16 0.82 When I feel happy, the moods of those around me improve 0.89 0.20 0.79 My happiness is contagious 0.84 0.14 0.71 My happiness doesn't influence the way others feel (rc) 0.60 0.27 0.38 My happiness spreads easily to others 0.81 0.08 0.68 My happiness doesn't have an effect on others' feelings (rc) 0.70 0.27 0.50 When I am sad, it doesn't affect the moods of those around me(rc) 0.16 0.64 0.40 My sadness worsens others' moods 0.14 0.77 0.60 When I feel sad, people around me are likely to become sad 0.12 0.65 0.43 My sadness affects the way that those around me feel 0.22 0.79 0.63 My sadness infects the people around me 0.30 0.80 0.64 Others are more likely to become sad if they are around me when I am sad 0.17 0.73 0.53 When I feel sad, the feeling is not easily spread to people around me (rc) 0.21 0.79 0.62 My sad feelings do not influence the moods of those around me (rc) 0.21 0.74 0.54 My sadness rubs off on other people 0.21 0.83 0.69 My sadness doesn't influence the way others feel (rc) 0.21 0.79 0.63

Table 2: Study 1B Demographics Student mTurk X2 / p - value ( n = 129) (n = 209) Age (SD) 19.81 (1.92) 34.64 (12.63) 13.23*** % Female 71 61 3.71 t Race % White 64 70 4.02 % Black 6 8 % Asian 10 10 % Hispanic 8 4 % Other/Multi 12 8 t p < .10; * p < .05; ** p < .01; *** p < .001

Table 3: Correlations between BASES scales and all variables of interest 1. 2. 3. 4. 5. 6. 7. 1. BASE-Happy 1.00 2. BASE-Sad .43 *** 1.00 *** *** 3. Emotion Contagion .33 .31 1.00 *** *** *** 4. Positive Affect .46 .19 .23 1.00 *** *** 5. Negative Affect -.24 -.03 .06 -.19 1.00 *** ** * *** * 6. Extraversion .27 .25 .20 .44 -.22 1.00 ** ** ** * 7. Agreeableness .23 .05 .26 .23 -.22 -.04 1.00 *** *** 8. Conscientiousness .06 .09 -.05 .34 -.33 .13 .13 * ** *** *** ** 9. Neuroticism -.17 .00 .24 -.34 .59 -.29 -.17 * ** 10. Openness .19 .09 .07 .10 -.14 .30 .11 11. Spread of positive .46 *** .29 *** .37 *** .30 *** -.14 * .01 .14 emotions to ingroup 12. Spread of negative .31 *** .42 *** .38 *** .20 *** .08 .12 .10 emotions to ingroup 13. Spread of positive .14 * .07 .17 ** .20 *** -.05 -.00 -.03 emotions to outgroup 14. Spread of negative .03 .12 * .16 ** .09 .16 ** -.01 .03 emotions to outgroup t p < .10; * p < .05; ** p < .01; *** p < .001

Table 3 (cont’d): Correlations between BASES scales and all variables of interest 8. 9. 10. 11. 12. 13. 14. 8. Conscientiousness 1.00 9. Neuroticism -.34 *** 1.00 10. Openness -.07 -.16 1.00 11. Spread of positive -.01 .11 .07 1.00 emotions to ingroup 12. Spread of negative -.07 .18 * .18 * .59 *** 1.00 emotions to ingroup 13. Spread of positive -.12 .04 -.17 .31 *** .17 ** 1.00 emotions to outgroup 14. Spread of negative -.20 * .09 -.05 .15 ** .30 *** .70 *** 1.00 emotions to outgroup t p < .10; * p < .05; ** p < .01; *** p < .001

Table 4a: Results of hierarchical linear models predicting reported emotional behaviors: Expressive suppression B SE β t ΔR 2 Step 1 0.21*** Age 0.00 0.01 -0.03 -0.50 Sex (Ref: Female) 0.34 0.13 0.13 2.61** PA -0.20 0.08 -0.13 -2.53* NA 0.17 0.08 0.11 2.09* Contagion -0.68 0.11 -0.34 -6.25*** Step 2 0.04*** Age 0.00 0.01 -0.03 -0.61 Sex (Ref: Female) 0.34 0.13 0.13 2.67** PA -0.11 0.08 -0.07 -1.30 NA 0.12 0.08 0.08 1.43 Contagion -0.51 0.11 -0.25 -4.47*** BASE-Hap -0.22 0.08 -0.17 -2.74** BASE-Sad -0.13 0.07 -0.11 -1.99* Step 3 0.003 ** <0.001 * Step 4 * t p < .10; * p < .05; ** p < .01; *** p < .001

Table 4b: Results of hierarchical linear models predicting reported emotional behaviors: Positive nonverbal expressivity B SE β t ΔR 2 Step 1 0.47*** Age 0.00 0.00 0.02 0.41 Sex (Ref: Female) -0.19 0.09 -0.09 -2.09* PA 0.40 0.05 0.32 7.39*** NA -0.03 0.06 -0.02 -0.47 Contagion 0.86 0.08 0.50 11.37*** Step 2 0.01* ** Age 0.00 0.00 0.03 0.60 Sex (Ref: Female) -0.19 0.09 -0.09 -2.09* PA 0.35 0.06 0.27 6.10*** NA 0.01 0.06 0.01 0.14 Contagion 0.79 0.08 0.46 9.84*** BASE-Hap 0.14 0.06 0.13 2.55** BASE-Sad 0.01 0.05 0.01 0.18 Step 3 0.01 *** Step 4 0.00 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 4c: Results of hierarchical linear models predicting reported emotional behaviors: Negative nonverbal expressivity B SE β t ΔR 2 Step 1 0.21*** Age 0.01 0.00 0.06 1.08 Sex (Ref: Female) -0.26 0.12 -0.11 -2.16* PA -0.06 0.07 -0.05 -0.89 NA 0.07 0.07 0.05 0.96 Contagion 0.74 0.10 0.40 7.45*** Step 2 0.02* Age 0.00 0.00 0.05 0.93 Sex (Ref: Female) -0.26 0.12 -0.11 -2.22* PA -0.08 0.08 -0.06 -1.07 NA 0.07 0.08 0.05 0.95 Contagion 0.65 0.11 0.35 6.19*** BASE-Hap 0.00 0.07 0.00 0.04 BASE-Sad 0.16 0.06 0.14 2.61** Step 3 0.02 Step 4 0.00 t p < .10; * p < .05; ** p < .01; *** p < .001

Table 4d: Results of hierarchical linear models predicting reported emotional behaviors: Capitalization B SE β t ΔR 2 Step 1 0.35*** Age 0.01 0.00 0.10 2.01* Sex (Ref: Female) -0.08 0.12 -0.03 -0.64 PA 0.41 0.07 0.27 5.70*** NA -0.22 0.07 -0.15 -2.99*** Contagion 0.81 0.10 0.40 8.17*** Step 2 0.03***

Age 0.01 0.00 0.10 2.10* Sex (Ref: Female) -0.08 0.12 -0.03 -0.67 PA 0.33 0.07 0.22 4.46*** NA -0.18 0.08 -0.12 -2.41* Contagion 0.66 0.10 0.33 6.34*** BASE-Hap 0.17 0.07 0.13 2.37* BASE-Sad 0.13 0.06 0.11 2.20* Step 3 0.01 *** Step 4 0.00 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 4e: Results of hierarchical linear models predicting reported emotional behaviors: Distress disclosure B SE β t ΔR 2 Step 1 0.24*** Age 0.00 0.00 0.00 -0.08 Sex (Ref: Female) -0.16 0.09 -0.09 -1.68 t PA 0.11 0.06 0.10 1.96* NA -0.13 0.06 -0.12 -2.32* Contagion 0.60 0.08 0.41 7.70*** Step 2 0.02**

Age 0.00 0.00 -0.02 -0.31 Sex (Ref: Female) -0.16 0.09 -0.09 -1.76 t PA 0.11 0.06 0.10 1.80 t NA -0.14 0.06 -0.13 -2.40* Contagion 0.53 0.08 0.36 6.45*** BASE-Hap -0.03 0.06 -0.03 -0.53 BASE-Sad 0.15 0.05 0.17 3.18*** Step 3 0.00 *** Step 4 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Chapter 3: Explaining Relationships between Beliefs about the Spread of Emotions

and Reported Emotional Behavior in the U.S. and South Korea

In Study 1, I investigated the nature of beliefs about the spread of emotions and examined the relationships between these beliefs and self-reported emotional behaviors. The data suggest beliefs about the spread of emotion are valence specific, with people tending to endorse the idea that their happiness spreads more strongly than the idea that their sadness spreads. There is some evidence, however, that these beliefs about the spread of happiness and sadness might generalize to beliefs about other positive and negative emotions, respectively. I explore this possibility further in the current study. Additionally, Study 1 finds that beliefs about the spread of one’s emotions are not merely a reflection of how readily individuals catch others’ emotions, nor are these beliefs a proxy for more general traits related to temperament.

Of primary interest to continued investigations, in Study 1, beliefs about the spread of happiness were associated with increased reported positive emotional behavior. That is, those that believed their happiness spread were less likely to report that they use expressive suppression and more likely to report that they express their positive emotions nonverbally and talk about their positive experiences with others. I suggested in the introduction that beliefs about the spread of emotions may relate to emotional behavior through a number of potential mechanisms, three of which may explain the relationships between beliefs about the spread of happiness and reported behavior. First, people may believe they express their happiness because they are motivated to benefit others. In addition, sharing positive affect with others is one important means of bonding and maintaining close relationships (Gable & Reis, 2010). People

41 may believe they express their happiness because they are motivated to strengthen their bonds with others. Finally, more self-focused social goals might explain this relationship. When others respond to our positive emotions by sharing our joy, their responses are perceived to be more supportive. People may believe they express their positivity because they are motivated to elicit high-quality support.

Perhaps more surprisingly, belief that one’s sadness spreads was also associated with increased reported emotional behavior in Study 1. Those that believed their sadness spread were less likely to report using expressive suppression, more likely to report expressing their emotions

(both positive and negative) nonverbally, and more likely to report talking about their positive and negative experiences with others. The patterns of results suggest that people who believe their sadness spreads may be less driven to make others feel good in the short-term—as spreading one’s sadness to them inherently thwarts this goal. Instead, as is the case with happiness, sharing negative emotions facilitates the formation and maintenance of close relationships (Collins & Miller, 1994; Glover & Parry, 2008; McPherson et al., 2001), and empathic responses to negative emotions are perceived to be more supportive (Dakof & Taylor,

1990; Han et al., 2011; Thoits, 2011; though see also Nils & Rimé, 2012). People who believe their sadness spreads may be more likely to endorse motivations to strengthen their relationships with others or to receive support from others.

Study 1 is limited in that it does not test these potential explanations for the relationships it finds. Study 2 addresses this limitation by including measures of these potential motivations along with measures of beliefs about the spread of emotions and perceived emotional behavior.

In addition, Study 2 incorporates data collected in South Korea. As described in the introduction,

Korea provides a theoretically meaningful contrast to the United States for the purposes of this study. People in East Asian contexts tend to report greater concern over the potential negative consequences of their positive emotions for others than do European Americans, meaning that spreading positive emotions to others may not be perceived as straightforwardly beneficial in this context (Lu & Gilmour, 2004; Uchida & Kitayama, 2009; Uchida et al., 2004). Further, Koreans in particular report being more concerned about burdening others with their negative feelings

(Kim et al., 2006; Taylor et al., 2004). Koreans who believe their sadness spreads may be more motivated to suppress their expression of negative emotions in pursuit of relational harmony.

Finally, the distress disclosure index used in Study 1 measures the tendency to discuss negative emotions, but it does not assess the range of more specific strategies participants may utilize in doing so. Study 2 measures three specific coping strategies in efforts to clarify the relationship between beliefs in the spread of sadness and interpersonal coping responses to negative life experiences. In particular, people’s tendencies to vent about their experiences (i.e., focus on their distress and let those emotions show), seek emotional support (i.e., seek comfort and understanding from others) and seek instrumental support (i.e., seek practical advice or tangible assistance) are assessed.

In summary, Study 2 addresses four key questions:

1. Is the structure of beliefs about the spread of emotions similar in the U.S. and

Korea?

2. Are relationships between beliefs about the spread of emotions and reported

behavior similar in the U.S. and Korea?

3. What strategies for discussion of negative emotions do people who believe

their sadness spreads report using?

4. Do interpersonal motives explain relationships between beliefs about the

spread of emotions and reported behavior?

To answer these questions, I conducted a survey of American and South Korean college students assessing beliefs about the spread of emotions, perceived emotional behaviors (including specific interpersonal coping strategies), and interpersonal motives.

Study 2

Method

Participants. Korean ( n = 105) and American ( n = 107) university students completed this study. Korean participants were recruited from a private university in a large Korean city and received monetary raffle prizes in exchange for their participation. Korean participants were, on average, 22.93 years old ( SD = 3.47) and were approximately 50% female ( n = 52). American students were recruited from a Jesuit university on the East Coast and received partial course credit in exchange for their participation. Seventeen American participants did not provide information on their gender, 20 on their age, and 27 on their race. Among those that did respond to these questions, participants were, on average, 20.68 years old ( SD = 5.67). Approximately

80% of participants reported being female ( n = 71). Eighty percent of American participants identified as being White ( n = 64), 10% as Asian ( n = 8), and 10% as being other or multiracial

(n = 8). American participants tended to be younger than Koreans ( t (189) = 3.37, p = .001) and were more likely to be female ( X2 (1, N = 194) = 17.35, p < .001). Because these demographic

44 data were missing from multiple participants, they were not included in the analyses reported below.

Measures.

Self-construal . Independent and interdependent self-construal was measured using the

Singelis (1994) scale. Twelve items measured independence (e.g., “I act the same way no matter who I am with”; α = .72 for Koreans, .78 for Americans) and 12 items measured interdependence

(e.g., “It is important for me to maintain harmony within my group”; α = .79 for Koreans, .80 for

Americans). Items were rated on a 7-point scale (1 = Strongly Disagree, 7 = Strongly Agree).

Koreans and Americans did not differ in their levels of reported independent self-construal ( t

(186) = 1.14, p = .26). However, American participants were higher in their ratings of interdependent self-construal than were Koreans ( t (186) = 5.34, p < .001). Past research has found that counter-intuitive findings patterns such as these are actually quite common in studies including self-reported self-construal (for a review, see Matsumoto, 1999; Oyserman, Coon, &

Kemmelmeier, 2002).

Beliefs about the spread of emotions . A shortened version of the BASE scale was used in this study, with the happiness and sadness subscales of the BASE scale reduced to 5-items each. This was done in efforts to minimize redundancy in some of the scale items, as well as to allow for the inclusion of items assessing beliefs about the spread of additional positive and negative emotions. Items selected to be retained, as well as additional items, are listed in Table 5.

Many items excluded from the revised scale emphasized emotional influence more broadly, without specifying the manner in which one’s emotions would influence others (e.g., “My sadness affects the way that those around me feel.”). Here, I restrained items to more specifically

45 assess beliefs about the spread of emotions—that is, beliefs that one’s happiness makes others feel happy, and one’s sadness makes them feel sad.

In order to further test whether beliefs about the spread of one’s happiness and sadness generalize to other positive and negative emotions, items assessing beliefs in the spread of other emotions were included. Two items each assessed beliefs in the spread of anger and excitement, and single items were used to assess beliefs in the spread of calmness, nervousness, guilt, embarrassment, and pride (see Table 5 for a full list of items). These latter three emotions were included to assess whether emotions with a more explicit interpersonal focus were conceptualized differently from the more “basic” emotions.

Dependent variables . Participants completed measures of nonverbal expressivity (BEQ; positive α = .73 for Koreans, .74 for Americans; negative α = .71 for Koreans, .76 for

Americans), and capitalization ( α = .88 for Koreans, .84 for Americans), as described in Study 1.

Additionally, three subscales of the COPE inventory (Carver, Scheier, & Weintraub, 1989), assessed participants’ tendencies to engage in different types of coping behaviors. Four items each measured participants’ tendencies to cope by venting (e.g., “I get upset and let my emotions out;” α = .68 for Koreans, .80 for Americans), emotional support seeking (e.g., “I get sympathy and understanding from someone;” α = .81 for Koreans, .86 for Americans), and instrumental support seeking (e.g., “I try to get advice from someone about what to do;” “I talk to someone who could do something concrete about the problem” α = .79 for Koreans, .78 for Americans).

Items were rated on a 4-point scale (1 = I usually don’t do this at all, 4 = I usually do this a lot).

Control variables . Participants completed the PANAS measure of trait affect as described in Study 1 (positive affect: α = .83 for Koreans, .88 for Americans; negative affect: α = .84 for

Koreans, .88 for Americans).

Interpersonal motives . A 36-item measure of interpersonal motives was adapted from the relational concerns scale used by Taylor et al. (2004) and the Circumplex Scale of Interpersonal

Values (CSIV; Locke, 2000). Items assessed the importance of a range of interpersonal goals when interacting with others. Participants were given the prompt, “ When I am with others, it is important that…” They were then asked to rate a series of motives on a 5-point scale (1 = not important, 5 = extremely important).

One subscale of this measure assessed the motive to benefit others (7-items; e.g.,

“…They do not become upset,” “…I make them feel good;” α = .78 for Koreans, .73 for

Americans). Another subscale assessed the motive to strengthen relationships (3-items; e.g., “…I feel close to them;” α = .73 for Koreans, .74 for Americans).a A third subscale assessed the motive to receive support (5-items; e.g., “…they comfort me,” “they understand how I’m feeling;” α = .86 for Koreans, .83 for Americans).

Results

Structure of beliefs in the U.S. and Korea. We first conducted exploratory factor analyses separately for Koreans and Americans to test the structure of beliefs about the spread of one’s emotions when emotions besides happiness and sadness were included in the scale. For

Americans, a three-factor solution was determined to be the best fit for the data. The first factor comprised items assessing beliefs about the spread of sadness and anger, and explained 23.77% a A fourth item (“…I avoid causing an argument”) was intended for inclusion in this scale. This item was excluded because reliability analyses indicated that it did not correlate well with other items, and its inclusion reduced reliability of the overall scale below an alpha of .70. 47 of the variance after rotation. Factor loadings for the first factor ranged from .46 to .81. The second factor, corresponding to the items measuring beliefs about the spread of happiness and excitement, explained an additional 17.75% of variance after rotation, and factor loadings ranged from .61 to .80. The third factor included items measuring beliefs about the spread of one’s guilt, embarrassment, and pride. This factor explained an additional 5.83% of variance and factor loadings ranged from .40 to .73. Belief in the spread of one’s calm and nervous feelings loaded weakly onto multiple factors.

For Koreans, a two-factor solution emerged as best fitting the data. The first factor reflected beliefs about the spread of happiness. This factor explained 28.29% of the variance after rotation, with factor loadings ranging from .40 to .84. The second factor, explaining an additional 8.45% of the variance, included beliefs about the spread of sadness (though two of these items also loaded with Factor 1), anger, and guilt. In addition, one of the items assessing beliefs about the spread of excitement loaded onto this factor. Beliefs about the spread of nervousness, embarrassment, pride, and calm did not load onto either factor, based on a minimum .40 factor loading.

As a result of these analyses, and to ease comparison between these results and the results of Study 1, only the items assessing beliefs about the spread of happiness and sadness were used in remaining analyses. These items form reliable scales in both cultural contexts. Cronbach’s alpha for the happiness subscale was .88 for Koreans and .83 for Americans. Cronbach’s alpha for the sadness subscale was .84 for Koreans and .87 for Americans b. While beliefs about the

b I also ran all analyses replacing the BASE-Sad scale with (1) only the three items from the sadness subscale that did not load onto both factors in the Korean factor solution ( α = .78 for Koreans, .83 for Americans), and (2) a scale reflecting beliefs about the spread of sadness and anger ( α = .82 for Koreans, .86 for Americans). Doing so did not alter the patterns of results from those reported in the text. 48 spread of anger loaded similarly for both groups, beliefs about other emotions loaded differently for Americans and Koreans.

We next ran two independent-samples t-tests to assess whether Koreans and Americans differed in mean levels of beliefs about whether their happiness and sadness spread (see Figure

1). Americans were more likely than Koreans to endorse the idea that their happiness spread ( t

(210) = 3.36, p = .001), but less likely to believe that their sadness spread ( t (210) = 3.09, p =

.002). Paired samples t-tests further demonstrated that both Koreans and Americans endorse the belief that their happiness spreads more than that their sadness spreads (t (104) = 5.37, p < .001 for Koreans; t (106) = 11.63, p < .001 for Americans). However, the difference between these beliefs was narrower among Koreans than Americans; Americans endorsed the belief that happiness spread more strongly than Koreans ( t (210) = 3.36, p = .001) and the belief that sadness spread less strongly than Koreans ( t (210) = 3.09, p = .002). Finally, in line with results of factor analyses, beliefs about the spread of happiness and beliefs about the spread of sadness were more strongly correlated among Koreans ( r (104) = .52, p < .001) than Americans ( r (106)

= .22, p = .03; z = 2.82, p = .002).

Relationships with self-reported emotional behavior. To test whether the relationships between beliefs in the spread of emotions and reported emotional behavior differed across cultural contexts, I ran a series of seven hierarchical multiple regression analyses predicting reports of expressive suppression, positive expressivity, negative expressivity, capitalization, venting, emotional support seeking, and instrumental support seeking. All dependent variables were strongly correlated for both Americans and Koreans, but were statistically distinct (see

Table 6). The first step of the regressions assessed main effects of cultural group, positive affect,

49 and negative affect. Beliefs about the spread of happiness and beliefs about the spread of sadness were entered in the second step to assess their contribution to the overall model. The third step included the interactions between culture and beliefs about the spread of happiness and culture and beliefs about the spread of sadness. A fourth step included the interaction between belief in the spread of happiness and belief in the spread of sadness.

As shown in Table 7, there was no significant main effect of culture in predicting most reports of emotional behavior, indicating that Americans and Koreans did not differ in their overall levels of perceived expressive suppression, nonverbal expressivity, capitalization, venting, or emotional support seeking (all p’s > .30). However, there was a main effect of culture in predicting instrumental social support seeking (β = .28, t (206) = 3.67, p < .001), such that

Koreans reported seeking instrumental support more often than did Americans.

Beliefs about the spread of happiness and sadness significantly contributed explanatory power beyond that provided by culture, positive affect, and negative affect in each of the regression models (see Table 7). Beliefs about the spread of happiness predicted greater reported positive nonverbal behavior (β = .45, t (206) = 6.36, p < .001), capitalization (β = .29, t (206) =

3.79, p < .001), emotional support seeking (β = .23, t (206) = 3.02, p = .003), and instrumental support seeking (β = .27, t (206) = 3.49, p = .001), but not negative nonverbal behavior (β = .05, t

(206) = .67, p = .50) or venting (β = -.01, t (206) = -.17, p = .87). Beliefs about the spread of sadness predicted greater reported negative nonverbal expressivity (β = .22, t (206) = 2.90, p =

.004), venting (β = .22, t (206) = 2.85, p = .005), and emotional support seeking (β = .16, t (206)

= 2.10, p = .037), but not positive nonverbal expressivity (β = .08, t (206) = 1.20, p = .23),

50 capitalization (β = .12, t (206) = 1.65, p = .10), or instrumental support seeking (β = .01, t (206) =

.14, p = .89).

Applying a Bonferroni family-wise correction for multiple tests, dividing a p-value of .05 by the number of tests (in this case, 7) results in an adjusted significance threshold of p = .007

(.05/7 = .007). The only finding affected by the application of this more conservative criteria is the relationship between beliefs about the spread of sadness and reported emotional support seeking, which was no longer statistically significant ( p = .04 > adjusted p = .007).

Interactions between culture and beliefs about the spread of emotions failed to add explanatory power to any of the regression models, suggesting that the main effects of beliefs about the spread of happiness and sadness were similar in both the U.S. and Korea (all ΔR 2 <

2%). The inclusion of the interaction between belief in the spread of happiness and belief in the spread of sadness also did not add explanatory power to these models.

Explaining relationships with reported behavior. We hypothesized that interpersonal motives would mediate the relationships reported above. To test this possibility, it was necessary to first test whether beliefs in the spread of happiness and sadness predicted motives to benefit others, to strengthen relationships, or to receive support, again controlling for culture, positive affect, and negative affect c.

As shown in Table 8, Koreans endorsed motivations to benefit others and to receive support from others more strongly than did Americans ( β = -.37, t (207) = -5.17, p < .001; β = -

.23, t (207) = -3.03, p = .003), but groups did not differ in their motivations to strengthen

c It would also have been possible for interpersonal motives to moderate the relationships between beliefs about the spread of emotions and emotional behavior. This possibility was tested, but the inclusion of interactions between beliefs about the spread of emotions and interpersonal motives did not significantly contribute to the explanatory power of the models predicting any of the dependent variables of interest. 51 relationship ( β = -.02, t (207) = -.29, p = .77). Beliefs about the spread of happiness and sadness significantly contributed to explanatory power in each of these three models. The belief that happiness spreads predicted motivations to strengthen relationships ( β = .19, t (207) = 2.47, p =

.01), but not to benefit others (β = -.01, t (207) = -.11, p = .91) or to receive support ( β = -.05, t

(207) = -.62, p = .53). The belief that sadness spreads predicted motives to benefit others (β =

.25, t (207) = 3.42, p = .001) and to receive support ( β = .33, t (207) = 4.58, p < .001) but not to strengthen relationships (β = .12, t (207) = 1.58, p = .12). Including interactions between culture and beliefs about the spread of happiness and sadness again did not significantly add explanatory power in predicting interpersonal motives (all ΔR 2 < 2%), indicating that the strength of relationships between beliefs and interpersonal motives were similar across cultures. Therefore, mediation analyses reported below are collapsed across cultures.

Beliefs about the spread of happiness. As reported above, belief in the spread of happiness predicted perceptions of expressive suppression, positive nonverbal expressivity, capitalization, emotional support seeking, and instrumental support seeking. This belief also predicted the motivation to strengthen relationships. I tested whether this motive mediated relationships with reported emotional behavior using a bootstrapping approach with resampling size of 5000 (Preacher & Hayes, 2008). As shown in Figure 2, the motive to strengthen relationships partially mediated relationships between belief in the spread of happiness and four of the five reported emotional behavior outcomes it was initially associated with (i.e., positive nonverbal behavior, capitalization, emotional support seeking, instrumental support seeking, but not suppression).

Beliefs about the spread of sadness. As reported above, belief in the spread of sadness predicted perceptions of suppression, negative nonverbal expressivity, venting, and emotional support seeking. It also predicted motives to make others feel good and to receive support. Again using a bootstrapping method with resampling size of 5000 (Preacher & Hayes, 2008), two separate analyses tested whether either the motives to benefit others or to receive support mediated relationships with reported behavior. The motivation to make others feel good did not significantly mediate any of these relationships (95% confidence intervals include zero).

However, as shown in Figure 3, there is evidence for mediation of the relationships between belief in the spread of sadness and each reported behavior (i.e., suppression, negative nonverbal expressivity, venting, emotional support seeking d) through the motivation to receive support.

Testing alternate causal pathways. Of course, these data are cross-sectional, meaning that the analyses above infer a directionality that cannot be empirically confirmed. It would theoretically be possible for alternative causal pathways to also explain these data. The motivations to strengthen relationships or to receive support from others may spur open expression of positive and negative emotions, respectively. This open expression may lead them to see evidence of the spread of emotions more frequently, causing the belief that emotions spread. To examine whether this alternative causal pathway fits the data, I ran a set of mediation models in which interpersonal goals predict beliefs about the spread of emotions through their relationships to emotional behavior.

Using this alternate causal pathway, there was evidence of mediation for models predicting belief in the spread of happiness from the goal to strengthen relationships through

d Recall that the relationship between beliefs about the spread of sadness and reported emotional support seeking was not statistically significant following Bonferroni adjustment. 53 reported positive nonverbal expressivity, capitalization, emotional support seeking, and instrumental support seeking (95% CI’s do not include 0), but not suppression (95% CI’s include

0). These results indicate that belief in the spread of happiness may encourage increased positive emotional behavior (as suggested by the originally tested pathway), and also result from increased positive emotional behavior (as suggested by the alternate pathway; see Table 9 for indirect effect sizes in each model). However, there was no evidence of mediation for models predicting beliefs about the spread of sadness from the motivation to receive support through reported emotional behaviors (95% CI’s include 0; see Table 10). These results indicate that belief in the spread of sadness may not be as strongly influenced by emotional behavior as belief in the spread of happiness e.

Discussion

In line with the results of Study 1, the structure of beliefs about the spread of emotions was largely valence-based for American participants. In addition, Americans do not think of more explicitly social emotions (i.e., guilt, embarrassment, and pride) spreading similarly to how they think of happiness or sadness spreading. Koreans did not share the same factor structure as

Americans—instead distinguishing beliefs about the spread of happiness from most other emotions. Still, beliefs about the spread of one’s happiness and one’s sadness could be considered separate constructs in both cultural contexts.

Replicating the results of Study 1, Study 2 finds that the more people believe their emotions spread, the more they report expressing their emotions. As in Study 1, the belief that

e Additional pathways, such as the indirect effect of motivations on behaviors through beliefs about the spread of emotions, were also tested. Though some of these did yield evidence of statistically significant mediations, for the sake of parsimony these are not discussed in detail. However, these should serve to emphasize our inability to infer causal relationships using the current data. 54 happiness spread was associated with decreased reported suppression and increased reported expression of positive emotions. In Study 1, the belief that sadness spread was associated with increases in both positive and negative reported emotional behavior, but in Study 2 this belief was associated only with reported expression of negative emotions. Further research is needed to understand the conditions under which belief in the spread of sadness might be associated with increased perceptions of one’s positive emotional behavior.

The use of the COPE scale in Study 2 clarified the particular coping behaviors that were associated with beliefs about the spread of emotions. The belief that happiness spreads was not associated with reports of general distress disclosure in Sudy 1, but it did predict increased reported emotional and instrumental social support seeking in Study 2. The belief that sadness spreads was related to reports of general distress disclosure in Study 1, and Study 2 clarified that believing sadness spreads predicts reported venting and emotional support seeking in response to negative life events, though the latter was statistically non-significant when correcting for family-wise error. It is possible that the measure of distress disclosure used in Study 1 was interpreted by participants to be more akin to venting than to more goal-directed interpersonal coping strategies like emotional and instrumental support seeking.

Another goal of Study 2 was to test whether the relationships observed in Study 1 varied across cultural contexts. Past research has indicated Koreans tend to be more concerned with the potential negative consequences of their emotions for others than are Americans. Thus, I hypothesized that belief in the spread of sadness may predict decreased, instead of increased, perceived expression of negative emotions in Korean context. Instead, Study 2 found no evidence of cultural variation in the relationships between beliefs in the spread of emotions and

55 reported emotional behaviors. In both the U.S. and South Korea, beliefs about the spread of happiness predicted increased reported positive emotional behavior and beliefs about the spread of sadness predicted increased reported negative emotional behavior.

Cultural variation between Western and East Asian contexts is often theoretically linked to differences in independent and interdependent self-construal. However, in this sample,

American and Korean participants did not differ in their reported levels of independent self- construal, and American participants endorsed interdependent self-construal more highly than did Koreans. It is possible that the measure used here did not accurately assess true levels of independence or interdependence; recent work has suggested these constructs may be more accurately measured via implicit or behavioral measures (Kitayama, Park, Timur, Karasawa, &

Uskul, 2009). It is also possible that reported differences in self-construal do accurately reflect these populations. Students from Jesuit universities have scored higher than other American college samples in measures of interdependent self-construal in past research, perhaps reflecting emphasis on Jesuit values of connectedness and social responsibility (Chentsova-Dutton & Tsai,

2010). University students and urbanites (like the Korean students in this sample) also tend to be more independently oriented than their non-student, rural compatriots (Henrich, Heine, &

Norenzayan, 2010). Therefore, it is possible that the particular characteristics of the American and Korean samples used in this research minimized cultural differences.

Finally, Study 2 examined the roles of other-focused and self-focused motives in explaining the relationships between beliefs and reported behaviors. The results suggest that the relatively other-focused goal to strengthen relationships best explains links between belief in the spread of happiness and reported positive expressivity. That is, the more individuals believe their

56 happiness spreads, the more important they say it is to feel close and connected to others, and this relationship at least partially explains their tendency to report expressing their emotions more. However, an alternate causal pathway may also explain these relationships: the more important it is for people to feel close to others, the more they report expressing their emotions, and the belief in the spread of happiness is explained by this increased expressivity. Belief in the spread of emotions, and particularly of happiness, may both encourage people to express their emotions in pursuit of interpersonal goals and also result from people’s tendency to express their emotions when such goals are present.

The relatively self-focused motivation to receive support best explains the links between belief in the spread of sadness and reported negative expressivity. It seems that the more individuals report believing their sadness spreads, the more important they believe it is that they receive understanding and support from others, and thus the less they suppress their emotions and the more they believe they express their negative emotions nonverbally, by venting, and by seeking emotional social support. The spread of sadness may be perceived as a useful means of eliciting empathy and understanding, best preparing others to provide the type of support people desire.

Limitations

One key limitation of the current research is its reliance on self-report. Across two studies, believing one’s emotions spread is associated with increases in perceived emotional behavior. However, perceptions of emotions do not map on precisely to emotions as they are actually experienced or expressed (Robinson & Clore, 2002). Future research would benefit from examining beliefs about the spread of emotions in the context of dynamic interpersonal

57 interactions. This would allow for testing whether belief in the spread of emotions is associated with actual, as well as perceived, increases in emotional behavior.

Additionally, throughout this paper I have discussed beliefs about the spread of emotions as predictors of reported behavior through interpersonal goal pursuit. However, this does not necessarily mean that beliefs about emotions cause these outcomes. A reverse causal pathway, in which the open expression of emotion or the presence of a particular interpersonal motive causes one to believe that emotions spread, is equally plausible. Indeed, mediation results suggest beliefs about the spread of happiness may be especially likely to result from emotional behavior.

There may also be some unmeasured variable that is associated with both beliefs about emotions and their outcomes that is a more valid or proximate cause of emotional behavior. Particular types of people (for example, extraverts) may be especially likely to possess motivations related to relationship building and support seeking, to express their emotions more, and to believe that emotions spread. Though I have tried to address these issues statistically, an experimental approach is needed to truly gauge the causal nature of these relationships.

Table 5: List of items included in the revised BASE scale Happiness 1. When I am happy, people around me become happier 2. My happiness brightens others' moods 3. The people around me become happier when I'm happy 4. When I feel happy, the moods of those around me improve 5. My happiness doesn't have an effect on others' feelings (rc) Sadness 6. My sadness worsens others' moods 7. When I feel sad, people around me are likely to become sad 8. My sadness infects the people around me 9. Others are more likely to become sad if they are around me when I am sad 10. When I feel sad, the feeling is not easily spread to people around me (rc) Calm 11. When I feel calm, people around me become calmer Excitement 12. When I am excited, others feel that excitement, too 13. Others will NOT feel more excited if they are around me when I am excited (RC) Anger 14. My angry feelings do NOT spread easily to others (RC) 15. Others become angrier when they are around me and I am angry Nervousness 16. When I am nervous, my nerves spread to those around me Guilt 17. When I feel guilty, the feeling is easily spread to others Embarrassment 18. When I am embarrassed, others are likely to catch that feeling Pride 19. When I am proud, the feeling spreads easily to others

Table 6: Correlations between all dependent variables among Koreans (lower half) and Americans (top half). 1. 2. 3. 4. 5. 6. 7. 1. Suppression 1 -0.60*** -0.59*** -0.55*** -0.58*** -0.39*** -0.68*** 2. Expressivity (pos) -0.60*** 1 0.55*** 0.59*** 0.59*** 0.51** 0.59*** 3. Expressivity (neg) -0.57*** 0.59*** 1 0.37*** 0.59*** 0.41*** 0.50*** 4. Capitalization -0.57*** 0.56*** 0.41*** 1 0.45*** 0.41*** 0.45*** 5. Venting -0.54*** 0.35*** 0.49*** 0.35*** 1 0.34*** 0.59*** 6. Emotional SS -0.38*** 0.29** * 0.29** * 0.43*** 0.45*** 1 0.54*** 7. Instrumental SS -0.54*** 0.39*** 0.39*** 0.54*** 0.59*** 0.64*** 1 t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7a: Results of hierarchical linear models predicting reported emotional behaviors: Expressive suppression B SE β t ΔR 2 Step 1 0.03 *** Culture -0.02 0.19 -0.01 -0.10 *** (Ref: Korean) PA -0.07 0.11 -0.04 -0.58 *** NA 0.23 0.10 0.15 2.20* ** Step 2 0.09*** Culture 0.00 0.18 0.00 0.01 *** (Ref: Korean) PA -0.01 0.11 -0.01 -0.12 *** NA 0.20 0.10 0.13 2.00* ** BASE-Hap -0.32 0.10 -0.23 -3.09*** BASE-Sad -0.14 0.08 -0.13 -1.78 t* * Step 3 0.01 *** Step 4 0.00 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7b: Results of hierarchical linear models predicting reported emotional behaviors: Positive nonverbal expressivity B SE β t ΔR 2 Step 1 0.03 *** Culture 0.00 0.15 0.00 0.01 *** (Ref: Korean) PA 0.20 0.09 0.17 2.27* ** NA -0.03 0.08 -0.02 -0.34 *** Step 2 0.21*** Culture -0.08 0.13 -0.04 -0.57 *** (Ref: Korean) PA 0.10 0.08 0.09 1.23 *** NA 0.03 0.07 0.02 0.40 *** BASE-Hap 0.48 0.08 0.45 6.36*** BASE-Sad 0.07 0.06 0.08 1.20 *** Step 3 0.02 *** Step 4 0.00 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7c: Results of hierarchical linear models predicting reported emotional behaviors: Negative nonverbal expressivity B SE β t ΔR 2 Step 1 0.03 *** Culture -0.20 0.16 -0.10 -1.30 ** (Ref: Korean) PA -0.01 0.09 -0.01 -0.11 ** NA 0.15 0.09 0.12 1.77 t * Step 2 0.06** * Culture -0.15 0.16 -0.07 -0.95 ** (Ref: Korean) PA 0.01 0.09 0.00 0.06 ** NA 0.13 0.09 0.11 1.55 ** BASE-Hap 0.06 0.09 0.05 0.67 ** BASE-Sad 0.20 0.07 0.22 2.90** Step 3 0.01 *** Step 4 0.00 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7d: Results of hierarchical linear models predicting reported emotional behaviors: Capitalization B SE β t ΔR 2 Step 1 0.02 *** Culture 0.13 0.18 0.05 0.71 *** (Ref: Korean) PA 0.07 0.11 0.05 0.65 *** NA -0.14 0.10 -0.10 -1.40 *** Step 2 0.11*** Culture 0.09 0.17 0.04 0.52 *** (Ref: Korean) PA 0.00 0.11 0.00 0.04 *** NA -0.10 0.09 -0.07 -1.09 *** BASE-Hap 0.37 0.10 0.29 3.79*** BASE-Sad 0.12 0.08 0.12 1.65 *** Step 3 0.00 *** Step 4 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7e: Results of hierarchical linear models predicting reported emotional behaviors: Venting B SE β t ΔR 2 Step 1 0.03 *** Culture -0.09 0.11 -0.06 -0.83 *** (Ref: Korean) PA -0.02 0.06 -0.02 -0.26 *** NA 0.13 0.06 0.16 2.31* ** Step 2 0.04** * Culture -0.04 0.11 -0.03 -0.39 *** (Ref: Korean) PA 0.01 0.06 0.01 0.10 *** NA 0.11 0.06 0.13 1.95* ** BASE-Hap -0.01 0.06 -0.01 -0.17 *** BASE-Sad 0.13 0.05 0.22 2.85*** Step 3 0.01 *** Step 4 0.00 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7f: Results of hierarchical linear models predicting reported emotional behaviors: Emotional support seeking B SE β t ΔR 2 Step 1 0.02 *** Culture -0.09 0.11 -0.06 -0.79 *** (Ref: Korean) PA 0.12 0.07 0.14 1.82 t* * NA -0.01 0.06 -0.01 -0.16 *** Step 2 0.10*** Culture -0.10 0.11 -0.06 -0.86 *** (Ref: Korean) PA 0.10 0.07 0.11 1.42 *** NA 0.00 0.06 0.00 0.04 *** BASE-Hap 0.19 0.06 0.23 3.02*** BASE-Sad 0.10 0.05 0.16 2.10* ** Step 3 0.02 *** Step 4 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 7g: Results of hierarchical linear models predicting reported emotional behaviors: Instrumental support seeking B SE β t ΔR 2 Step 1 0.02 *** Culture -0.34 0.11 -0.25 -3.26*** (Ref: Korean) PA 0.14 0.06 0.17 2.19* ** NA 0.01 0.06 0.01 0.13 *** Step 2 0.10*** Culture -0.38 0.10 -0.28 -3.67*** (Ref: Korean) PA 0.09 0.06 0.11 1.44 *** NA 0.04 0.06 0.04 0.62 *** BASE-Hap 0.20 0.06 0.26 3.49*** BASE-Sad 0.01 0.05 0.01 0.14 *** Step 3 0.02 *** Step 4 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 8a: Results of hierarchical linear models predicting interpersonal motives: Benefit others B SE β t ΔR 2 Step 1 0.15***

Culture -0.46 0.09 -0.37 -5.17*** (Ref: Korean) PA 0.24 0.05 0.32 4.46***

NA 0.10 0.05 0.13 2.09* **

Step 2 0.04** *

Culture -0.44 0.09 -0.35 -4.86*** (Ref: Korean) PA 0.25 0.05 0.33 4.50***

NA 0.09 0.05 0.12 1.90 t* *

BASE-Hap 0.04 0.05 0.05 0.75 ***

BASE-Sad 0.10 0.04 0.18 2.59** *

Step 3 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 8b: Results of hierarchical linear models predicting interpersonal motives: Strengthen relationships B SE β t ΔR 2 Step 1 0.04* **

Culture -0.03 0.11 -0.02 -0.29 *** (Ref: Korean) PA 0.16 0.06 0.19 2.51** *

NA 0.06 0.06 0.08 1.11 ***

Step 2 0.06***

Culture -0.04 0.10 -0.03 -0.37 *** (Ref: Korean) PA 0.14 0.06 0.16 2.11* **

NA 0.08 0.06 0.09 1.31 ***

BASE-Hap 0.15 0.06 0.19 2.49** *

BASE-Sad 0.07 0.05 0.11 1.48 ***

Step 3 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 8c: Results of hierarchical linear models predicting interpersonal motives: Receive support B SE β t ΔR 2 Step 1 0.03 *** Culture -0.36 0.12 -0.23 -3.03*** (Ref: Korean) PA 0.13 0.07 0.13 1.79 t* * NA 0.17 0.07 0.17 2.55** * Step 2 0.09***

Culture -0.29 0.12 -0.18 -2.46** * (Ref: Korean) PA 0.16 0.07 0.17 2.27* ** NA 0.14 0.06 0.14 2.11* ** BASE-Hap -0.01 0.07 -0.01 -0.18 *** BASE-Sad 0.21 0.05 0.30 4.07*** Step 3 0.01 *** t p < .10; * p < .05; ** p < .01; *** p < .001

Table 9: Indirect effects and bias-corrected confidence intervals of IV on DV through proposed mediators for belief in the spread of happiness and motive to strengthen relationships IV = Belief IV = Motive DV(s) = Reported behavior DV = Belief Mediator = Motive Mediator(s) = Reported behavior Indirect DV/Mediator Indirect Effect 95% CI Effect 95% CI

Suppress 0.01 [-.07 - .02] .02 [-.01 - .08]

BEQ-Pos 0.04 [.01 - .09] 0.11 [.04 - .21]

Capitalization 0.07 [.01 - .15] 0.09 [.03 - .19] Emotional Support Seeking 0.04 [.01 - .09] 0.06 [.01 - .12] Instrumental Support Seeking 0.03 [.01 - .08] 0.05 [.01 - .11] t p < .10; * p < .05; ** p < .01; *** p < .001

Table 10: Indirect effects and bias-corrected confidence intervals of IV on DV through proposed mediators for belief in the spread of sadness and motive to receive support IV = Belief IV = Motive DV(s) = Reported behavior DV = Belief Mediator = Motive Mediator(s) = Reported behavior Indirect DV/Mediator Indirect Effect 95% CI Effect 95% CI

Suppression -0.06 [-.02 - -.13] 0.02 [-.01 - .06]

BEQ-Neg 0.09 [.04 - .15] 0.05 [.00 - .14]

Venting 0.07 [.04 - .12] 0.05 [-.01 - .12] Emotional Support Seeking 0.07 [.04 - .13] 0.03 [-.04 - .09] t p < .10; * p < .05; ** p < .01; *** p < .001

4 Korean

Emotions American 3

2 Beliefs About the Spread Spread About the of Beliefs

1 BASE-Hap BASE-Sad

Figure 1: Mean levels of beliefs about the spread of happiness and sadness among Koreans and Americans

Figure 2: Mediation model displaying effects of belief in the spread of happiness on reported behaviors through the motive to feel close to others.

Figure 3: Mediation model displaying effects of belief in the spread of sadness on reported behaviors through the motive to receive support.

Chapter 4: Manipulating Beliefs about the Spread of Emotions: An Investigation

Testing the Causal Relationship between Beliefs and Observable Behavior

Across two studies and two cultures, those who believed their emotions spread reported expressing their emotions more. This is a novel addition to prior work on the implications of beliefs about emotions. I hypothesized that these relationships would be explained by interpersonal motives—particularly the relatively other-focused motivation to benefit others or to strengthen relationships, or the relatively self-focused motive to receive support. In line with this hypothesis, beliefs about the spread of happiness were associated with increased positive emotional behavior through the motivation to feel close to others. Beliefs about the spread of sadness were associated with increased negative emotional behavior through the motivation to receive support.

These initial studies were limited in a number of important ways, which are addressed in

Study 3. First, I address the issue of causality by using an experimental design in which beliefs about the spread of emotions are manipulated. Next, I address the potential discrepancy between how people perceive their own behavior and how they actually behave by assessing their emotional behavior as they discuss recent positive and negative emotional experiences. I expect the belief that emotions spread will be causally related to both positive and negative emotion outcomes. More specifically, based on the results of Study 2, I expect those led to believe their emotions spread to report trying to suppress their emotions less, to express their positive emotions more when discussing positive experiences, and to express their negative emotions more when discussing negative experiences. However, it is also possible that thinking about the spread of emotions in abstract terms yields different associations with reported behavior than it

72 would if people were confronted more concretely with the consequences of spreading emotions.

If people are forced to think through the actual ramifications of spreading emotions at a more conscious level, their behavior might be more strongly motivated by concerns about others— leading to decreased negative emotional behavior and increased efforts to regulate emotions.

Next, I have connected beliefs about the spread of emotions to reported behavior through interpersonal motivations. From this, I inferred that the spread of emotions may be perceived as an effective tool for achieving social goals. However, I have not yet tested this hypothesis explicitly. I do so in the current study by assessing whether beliefs about the spread of emotions are associated with the quality of support people expect to receive. I expect those led to believe their emotions spread will expect higher quality support, indicating the spread of emotions is perceived as an effective means of eliciting high-quality support.

Finally, Study 2 found that interpersonal motives explained the relationships between beliefs about the spread of emotions and reported behavior. However, the use of cross-sectional data limited the ability to rigorously test the directions of relationships among these variables.

The next step in this program of research is to test whether beliefs about the spread of emotions can be causally connected to emotional behavior. Using an experimental design, I can test whether manipulating beliefs influences participants’ endorsement of interpersonal motivations, and whether these in turn explain any changes in affective outcomes.

In summary, Study 3 addresses three key questions that remain after Studies 1-2:

1. Are beliefs about the spread of emotions causally related to observable behavior and

emotion regulation?

2. Does belief in the spread of emotions cause people to expect higher quality social

support?

3. Do interpersonal motives explain relationships between beliefs and emotion outcomes?

Study 3

Method

Participants. Ninety-seven Korean students at a university in Seoul, South Korea participated in this study for payment. Approximately half (47%) of Korean students identified as female, and participants were, on average, 22.27 years old ( SD = 2.72). One hundred and three students at a Jesuit University on the East coast participated in this study for partial course credit.

However, a total of 12 students were excluded from the final sample. Two students were excluded due to concerns about their English proficiency—these concerns were noted by experimenters and confirmed by assessing their English fluency in their video recordings. One participant was excluded due to computer error. Nine participants were excluded because they guessed the true nature of the experiment f. Participants were thoroughly debriefed following the experiment, at which time they were asked what they believed was the nature of the experiment.

Participants were excluded only if they a) expressed doubt regarding the stated purpose of the experimental prime and b) correctly identified the true purpose of the experimental primes. The final sample of American participants included 91 students. Of these, 79% ( n = 72) identified as

White/Caucasian, 8% (n = 7) as Asian, 4% ( n = 4) as Latino/Hispanic, and 4% ( n = 4) as

Black/African American, and 3% ( n = 3) as another or mixed ethnicity. The majority of

f No participants in Korea correctly guessed the manipulation. This may be due to differences in the subject pools between the U.S. and South Korea: participants in the U.S. were recruited from psychology courses, and many of these were recruited from social psychology courses, where the use of deception in experimental studies was explicitly taught. Korean students were recruited from the University as a whole, and therefore were less likely to have a background in social psychological research. 74

American participants identified as female (76%) and participants were, on average, 19.34 years old (SD = 1.16). American participants were significantly younger than Korean participants ( t

(173) = 8.89, p < .001) and significantly more likely to be female ( X2 (1, N = 175) = 14.26, p <

.001).

Procedure. Before arriving in the lab, participants completed a pre-laboratory online survey in which they first identified a close friend and then described three recent positive experiences and three recent negative experiences. Participants were told their experiences should not directly involve the friend they had identified, and also that they should only describe experiences they would feel comfortable discussing in a laboratory session. A trained research assistant selected one positive experience and one negative experience for each participant based on the reported timing of the event, with the most recent event selected for discussion. If events were equally recent, the experience rated as more intense was selected. The pre-laboratory survey also assessed self-construal as well as other cultural values and beliefs about emotions

(i.e., controllability, emotion control values, ideal affect) that are outside the scope of the current research.

Two Korean research assistants ran all participants in South Korea, and three American research assistants ran all participants in the U.S. Participants were randomly assigned to one of three conditions: Emotion Spread, No Spread, and Control. Each participant read an article that was created for this study. Articles described the fabricated results of psychological research, and were made to look as though they had been printed from a news website. Participants in the emotion spread condition read an article describing research demonstrating that emotions spread easily from person to person and through social networks. Participants in the “no spread”

75 condition read an article suggesting that emotions do not spread very easily through social networks. Participants in the “control” condition read an article about emotional development in infancy. All articles were created to be the same length and the same level of difficulty (see

Appendix). For all participants, the research described in the article was presented as part of the

Framingham Heart Study, conducted in the United States by researchers at Harvard University.

A description of research conducted in the U.S. may have influenced Koreans and Americans differently. More specifically, Koreans may have been less likely than Americans to believe that the results of the study were applicable to them, making this a more conservative test of the impacts of the manipulation in this population.

So students would not know the experimenter was attempting to influence their beliefs, the experimenter presented the articles as a separate study. Participants were asked if they were willing to participate in a short pilot study investigating people’s memories for information about psychology based on the type of print source (i.e., textbook, news article, etc.). All participants agreed, and then read the article they were assigned. After reading, participants were asked to proceed to the main study.

First, participants completed a short online survey asking them about their current affect.

All participants discussed both a positive and a negative experience. The order of these discussions was counterbalanced across participants, so that half of the participants discussed a positive experience first, and half discussed a negative experience first. The experimenter reminded participants about the friend they had identified in their online survey and asked them to describe the emotional experience selected for the first discussion (see selection criteria above). Participants were asked to imagine their friend was sitting in a chair positioned across

76 from them, and to act as if they were having a discussion with their friend. This method was chosen because past research demonstrates imagined social presence can have similar effects on emotional behavior as real social presence (Jakobs, Manstead, & Fischer, 1999). To acquaint participants with the task of speaking aloud, they were asked to begin by talking about their day before beginning the discussion of their emotional experience. At this point, the experimenter left the room, leaving the participant alone with instructions to contact the experimenter when they completed their discussion.

After their first discussion with their imagined partner, participants answered a series of questions about their feelings, motivations, and expectations during the discussion, followed by the manipulation check. Next, the experimenter reminded the participant of the other emotional experience they were to discuss, reiterating the instructions before leaving the room. Again, participants answered a series of questions about their feelings, motivations, and expectations following their second discussion. Participants were videotaped throughout each of the two discussions, and were thoroughly debriefed at the conclusion of the study. The experimenter asked what they believed was the purpose of the study, then explained the deception that had been used and the true design of the experiment.

This protocol was extensively piloted and adjusted based on feedback from participants and collaborators in South Korea. In particular, the article used in the “no spread” condition underwent numerous drafts. An early version asserted that researchers had found emotions did not spread to others at all, but pilot participants expressed strong skepticism to the article. As a result, this prime was adjusted so the effect of spreading emotions was described as weaker than expected, rather than nonexistent. Another change based on piloting was the inclusion of a period

77 of small-talk before beginning the emotional discussion. Pilot participants reported feeling uncomfortable speaking for the first minute or two of the discussion task, but said they soon began to feel more natural. A short baseline discussion was added so this initial discomfort might be less likely to impact data.

Finally, all measures used in the study as well as the script for interacting with participants underwent rigorous translation procedures to ensure linguistic equivalence. Materials were first translated from English into Korean by a bilingual collaborator fluent in both English and Korean. Another bilingual student then translated all materials from Korean back into

English and any discrepancies were discussed among team members. This process of translation and back-translation can yield a final product that, while accurate in literal translation, still differs in meaning across languages. Therefore, resulting Korean materials were subsequently reviewed by two members of the research team in South Korea, who identified any areas of the study materials or protocol that seemed confusing or culturally inappropriate. For example, the term “small-talk,” used in the English language script for experimenters’ interactions with participants, was not meaningful in the Korean context. Instead, participants were asked to describe their day.

Measures .

Self-construal. Independent and interdependent self-construal was measured in the pre- lab survey using the Singelis (1994) scale. Twelve items measured independence (e.g., “I act the same way no matter who I am with”; α = .45 for Koreans, .75 for Americans) and 12 items measured interdependence (e.g., “It is important for me to maintain harmony within my group”;

α = .59 for Koreans, .75 for Americans). Items were rated on a 7-point scale (1 = Strongly

Disagree, 7 = Strongly Agree). Americans reported higher levels of independent self-construal than Koreans ( t (186) = 4.46, p < .001), but also higher levels of interdependent self-construal than Koreans ( t (186) = 2.53, p = .012). This pattern is consistent with research showing that

Americans score higher than Koreans in explicit measures of individualism but that Koreans are often no higher than Americans in ratings of collectivism (Oyserman et al., 2002). This latter point is also consistent with the results of Study 2. However, it should be noted that due to the low reliability of this measure among Korean participants, results of analyses using this scale should be interpreted cautiously.

Manipulation checks. Participants answered eleven multiple choice questions about the article. Of these, six questions tested participants’ understanding of the main ideas of the article.

Other questions asked about more detailed content (i.e., sample size) and aspects of the article not related to its content (i.e., advertisements), so were not analyzed. Participants also reported how difficult, interesting, and believable they felt the article was. Finally, a single item asked participants the extent to which they agreed with the statement, “I believe emotions spread easily to others.”

Qualities of emotional experiences. When participants described their emotional experiences in the pre-laboratory survey, they also reported the intensity of the emotional experience (1 = Not at all intense, 5 = Extremely intense), when the experience occurred (1 =

Past three days, 5 = Past year), and whether the participant had already discussed the experience with their friend (1 = Yes, 2 = No).

Emotional experience. A baseline measure of positive affect (4-items) and negative affect (6-items) was collected immediately after participants read the experimental manipulation

79 article. These formed reliable scales among Koreans (positive: α = .89, negative: α = .82) and

Americans (positive: α = .83, negative: α = .68). Though Americans’ ratings of negative emotions did not form as reliable a scale as Koreans’, this is still an acceptable range.

Following each discussion, 8 items assessed the degree to which participants had felt each of four positive emotions (happiness, pride, affection, gratitude) and four negative emotions

(embarrassment, sadness, anger, and guilt) while discussing their experience. These items were not combined to form scales.

Nonverbal behavior. Nonverbal behavior was coded using the Emotional Expressive

Behavior (EEB) coding system (Gross, 1996). This is a global coding system, integrating descriptions of and judgments about behavior. It is meant for rating single subjects watching videos, and was adapted for use while participants were talking. For each emotion, the EEB system provides a 2-dimensional rating taking into account both the intensity and the duration of the emotion. Expressions can be rated as slight, moderate, or strong in intensity and as short or long in duration. A behavior is considered long in duration if it occurs for at least five seconds in a single occasion or is expressed on at least three separate occasions.

The EEB coding system includes guidelines for rating anger, confusion, disgust, fear, sadness, happiness, interest, and surprise. This list was reduced to focus on indicators of one positive emotion (happiness) and two negative emotions (anger, sadness). Confusion, interest, and surprise were excluded because there were no a-priori hypotheses regarding how these would be influenced by either experimental condition or culture. Disgust and fear were further excluded because these appeared at very low frequencies in a preliminary coding of

80 approximately ten percent of videos, making it difficult to obtain reliable codes of these facial behaviors across raters.

Happiness was indicated by smiling and laughter, with the intensity of happiness dependent upon the size of the smile, the presence of wrinkling around the eyes, and the presence of laughter. Anger was indicated by expressions such as furrowed eyebrows and pressed lips.

Behaviors such as head shaking and eye rolling were also indicative of anger. The intensity rating for anger was increased if these behaviors were expressed strongly or appeared together.

Sadness was indicated by raising the inner eyebrows and frowning, as well as behaviors such as sighing. Like anger, the rated intensity of sadness was increased if these behaviors were expressed strongly or appeared together. I also included a measure of embarrassment/discomfort, which primarily measured whether participants touched or covered their faces. For each emotion, coders provided a rating of the objective presence of the emotion based on the criteria above. In order to assess the degree to which objective codes match raters’ subjective impressions of emotional behavior, coders also provided a subjective rating of each emotion. Objective and subjective ratings were closely related for all emotions of interest: anger (positive discussion: r

(187) = .89; negative discussion: r (187) = .97), sadness (positive discussion: r (187) = .75; negative discussion: r (187) = .92), and happiness (positive discussion: r (187) = .95; negative discussion: r (187) = .95), objective measures are used in all analyses.

Coders were blind to which experimental conditions each participant was assigned.

Coders were also blind to the order in which participants discussed experiences, so they did not know whether each video was a discussion of a positive or a negative experience. Nonverbal behavior was coded with no sound, so that verbal behavior would not influence ratings. Further,

81 the same participant’s videos were never viewed back-to-back, so that coders could less easily infer the topic of the emotional discussion. Coders met frequently when beginning behavioral coding and routinely throughout the coding process. Ten percent of cases (n = 38) were coded by both coders and average intraclass correlations were calculated for each variable of interest.

Interrater reliability was adequate for ratings of anger (ICC = .83), sadness (ICC = .79), happiness (ICC = .77), and embarrassment (ICC = .86).

This rating system also includes an overall evaluation of the pleasantness (0 = Very unpleasant, 4 = Very pleasant) and expressivity (0 = Not expressive, 6 = Extremely expressive) in each video. In rating overall pleasantness and intensity, coders were advised to consider both peak and modal emotional expressions. These were also rated reliably by the two coders

(pleasantness: ICC = .73; expressivity: ICC = .69). However, both these ratings were excluded from analyses in favor of the emotion specific measures.

Verbal behavior. Linguistic Inquiry and Word Count (LIWC; Pennebaker, Booth, &

Francis, 2007) software was used to assess verbal emotional behavior from transcriptions of participants’ positive and negative discussions. This software provides a count of the positive and negative emotion words used during each discussion. A Korean version of LIWC has been created and used in past research. However, Dr. James Pennebaker, creator of the LIWC program suggested to me that more accurate measures of Koreans’ emotion word use can be obtained by running Korean text through google translate and running these English texts through the main

LIWC dictionary (personal communication, May 2015). Thus, this method was employed in the current analyses.

Emotion regulation. The Strategies Questionnaire (Ehring, Tuschen-Caffier, Schnülle,

Fischer, & Gross, 2010) was used to assess participants’ employment of emotion regulation strategies during each emotional discussion. Two items assessed efforts to suppress emotions (“I tried not to let my feelings show”; “I tried to suppress my emotions”). Items were rated on a 7- point scale (1 = Extremely disagree, 7 = Extremely agree). This measure was collected following both positive and negative discussions, but items from both discussions formed a single reliable scale, so reported efforts to suppress were collapsed across discussions ( α = .78 for Koreans, α =

.74 for Americans).

This scale also measured efforts to reappraise emotions. As in previous studies, this measure was not analyzed because it is less directly related to emotional behavior. Participants also completed a modified version of this scale assessing their perceptions of how well they had succeeded in suppressing and reappraising their emotions during each discussion. However, neither Koreans nor Americans seemed to distinguish strongly between their efforts to regulate and their success in doing so (.74 < r’s < .91). Therefore, these items were not analyzed further.

Expected support. Five items assessed expectations of support that the friend would provide (e.g., “My friend will want to support me”). Items were rated on a 7-point scale (1 =

Strongly disagree, 7 = Strongly agree). This measure was collected following both positive and negative discussions, but items from both discussions formed a single reliable scale so ratings of expected support were collapsed across discussions ( α = .88 for Koreans, α = 78 for Americans).

Interpersonal motives. A 36-item measure of interpersonal motives used in Study 2 was adapted for use in the current study. Items assessed the importance of a range of interpersonal goals when interacting with others. Though each motive was assessed following both positive

83 and negative discussions, these items also formed reliable scales when collapsed across discussions. After discussing each experience, participants were given the prompt, “ I discussed my experience the way that I did…” They were then asked to rate a series of motives on a 5- point scale (1 = Strongly disagree, 7 = Strongly agree). Of these, seven items assessed other- oriented motives to elicit positive or avoid causing negative emotions (e.g., “…So that my friend would not become upset,” “…to make my friend feel good;” α = .91 for Koreans, .88 for

Americans). Four items assessed the motive to strengthen relationships (“…So that I don’t damage my relationship with my friend,” “…because I want to feel close to my friend;” α = .79 for Koreans, .77 for Americans). Five items assessed the relatively self-focused motive to receive support (e.g., “…because I want my friend to comfort me,” “because I want my friend to understand how I’m feeling;” α = .84 for Koreans, .81 for Americans).

Additional measures. In addition to the measures described above, surveys also assessed participants’ perceptions of their physiological responding at baseline and during each emotional experience. These items were intended for a separate study on cultural differences in somatic responding during emotional events. A measure of participants’ efforts to express a range of emotions was also included. However, in most cases participants did not distinguish between the emotions they experienced and those they tried to express; thus, this measure did not provide sufficient range for inclusion in statistical analyses.

I also assessed how much participants expected and wanted their friend to feel a range of emotions. Participants’ expectation of their friend’s responding was conceptualized as a measure of anticipated spread of emotions, while their desired friend’s responding was conceptualized as a measure of the ideal level of emotion spread. I hypothesized that being led to believe emotions

84 spread might only be associated with increased emotional behavior if the friend were expected to

“catch” emotions about as strongly as the participant ideally wanted. If a friend was expected to

“catch” emotions more strongly than the participant believed was desirable, the participant might express emotions less if led to believe emotions spread. However, these items were strongly correlated with each other in the current data, indicating that participants largely believed their friend would feel the way they wanted them to. This may be an artifact of the study design: participants were told that they were to imagine talking about their experiences with a particular friend, so they may have been more likely to select friends that they know tend to respond in a desirable manner. Furthermore, these items do not clearly distinguish whether participants are conceptualizing their friends’ emotions as caused by the spread of emotions versus through other means (e.g., empathic responding), making any associations with these variables difficult to interpret. They are therefore not discussed in the current paper.

Results

Manipulation checks. Participants answered, on average, 88% of questions about the article correctly, indicating they paid attention to its content ( SD = 15%), and accurace did not differ across experimental conditions ( F (2, 184) = .81, p = .45). To examine the equivalence of experimental primes, one-way Analyses of Variance (ANOVA’s) tested whether participants across the three conditions found the articles equally difficult, interesting, and believable. To test the effectiveness of the experimental primes in altering beliefs about the spread of emotions, a one-way ANOVA tested responses to the item measuring belief in the spread of emotions.

Results of these analyses revealed there were no differences across conditions in participants’ perceptions of how difficult the article was ( F (2, 183) = .29, p = .75) or how

85 interesting it was ( F (2, 183) = 1.63, p = .20). However, there was an effect of condition on how believable people thought the article was ( F (2, 183) = 27.02, p < .001). Bonferroni-adjusted post-hoc mean comparisons indicated participants in the no spread condition were less likely to believe the content of their article than participants in the emotion spread or control conditions

(p’s < .001). This is consistent with the observation based on piloting that people tend to report this belief is counterintuitive. There was an effect of condition on the item assessing belief in the spread of emotions ( F (2, 183) = 6.45, p = .002), with participants in the no spread condition endorsing the belief that emotions spread less strongly than those in the emotion spread condition

(p = .009) or control condition ( p = .004). Participants in the emotion spread and control condition did not differ in their endorsement of this item ( p = 1.00). This indicates the manipulation resulted in some shift in beliefs about the spread of emotions g.

Qualities of emotional experiences. On average, emotional experiences selected for use in this study were reported as somewhat to very intense (Positive: M = 3.20, SD = 0.96;

Negative: M = 3.50, SD = .97), and occurred within a week of their laboratory visit (Positive: M

= 1.74, SD = 0.96; Negative: M = 1.94, SD = 1.02). Positive experiences had already been discussed with friends 57% of the time, and negative experiences 59% of the time. The equivalence of intensity and timing of emotional experiences across conditions and cultures was tested using four 2 (Culture) X 3 (Condition) ANOVA’s. Chi Square analyses tested whether likelihood of having shared the experience varied across conditions or cultures.

Across conditions, emotional experiences were equally intense (Positive: F (1, 181) =

0.50, p = .61; Negative: F (1, 181) = 0.23, p = .80), recent (Positive: F (1, 181) = 1.35, p = .26;

g Excluding participants who report strongly disbelieving the article results in a loss of power that reduces the size of some effects to marginal significance, but does not change the overall patterns of results from those reported below. 86

Negative: F (1, 181) = 0.04, p = .97), and likely to have been shared already (Positive: X2 (2, N =

186) = 0.48, p = .79; negative: X 2 (2, N = 187) = 1.99, p = .37). There were also no differences across cultures in the reported intensity of emotional experiences (Positive: F (1, 181) = 1.23, p =

.27; Negative: F (1, 181) = 0.63, p = .43) or the timing of positive experiences ( F (1, 181) = 0.98. p = .32). However, Korean participants tended to report more recent negative experiences than

Americans ( F (1, 181) = 14.60, p < .001). Additionally, while Koreans were equally likely to have already discussed their negative experiences (X2 (1, N = 187) = 2.20, p = .14), they were less likely than Americans to have already shared their positive experiences with their friend ( X2

(1, N = 186) = 10.07, p = .002). The inclusion of these variables as covariates in the models presented below did not meaningfully alter results and they did not significantly predict variables of interest. There were no significant interactions between culture and condition in predicting the intensity (Positive: F (1, 181) = 0.26, p = .77; Negative: F (1, 181) = 0.10, p = .90), or timing

(Positive: F (1, 181) = 0.08, p = .93; Negative: F (1, 181) = 0.65, p = .52) of reported experiences.

Emotional experience. Two 2 (Culture) X 3 (Condition) Analyses of Variance

(ANOVA’s) tested for effects of culture (Korean, American) and condition (Emotion spread, No spread, Control) on baseline positive and negative affect. There was no effect of condition on baseline positive affect ( F (2, 182) = 1.28, p = .28) or negative affect ( F (2, 182) = 1.97, p = .14).

However, main effects of culture revealed that Koreans reported experiencing significantly more positive affect at baseline ( F (1, 182) = 12.47, p = .001) and marginally more negative affect at baseline ( F (1, 182) = 3.36, p = .07). There was no significant interaction of condition and

87 culture in predicting positive affect ( F (2, 182) = 0.35, p = .71) or negative affect ( F (2, 182) =

1.55, p = .22).

Two 2 (Culture) X 3 (Condition) multivariate analyses of covariance (MANOVA’s) tested for effects of culture and condition on (1) positive emotions during positive discussions,

(2) negative emotions during negative discussions. Negative emotions were, on average, experienced at only very low rates during positive discussions, and positive emotions were experienced at low rates during negative discussions. There was not sufficient range in these variables to reliably analyze these cross-valenced emotions.

The MANOVA was chosen as the analytic method because it helps protect against the chance of Type 1 error that might occur if multiple conceptually related dependent variables are run in separate ANOVA’s. Each dependent variable included in the MANOVA is associated with loss of one degree of freedom. It is important to these analyses that dependent variables (1) appropriately account for outliers and (2) not be too strongly correlated with each other. Ratings of emotional experience did not contain outliers, and correlations between specific emotions were weakly to moderately related (Positive: .22 < r (187) < .55; Negative: .07 < r (188) < .38), signifying that it is appropriate to evaluate these variables together.

Positive discussions. The results of the MANOVA indicate a significant overall main

2 effect of culture ( λ = .92, F (4, 178) = 3.84. p = .005, ηp = .00). There was no significant overall effect of condition on positive emotions ( λ = .97, F (8, 356) = 0.81. p = .60), nor was there a significant interaction between culture and condition ( λ = 1.00, F (8, 356) = 0.09. p > .99).

Univariate analyses of variance (ANOVA) results helped to determine whether the effects of culture were constrained to particular variables. This effect was driven by cultural differences in

2 ratings of affection ( F (1, 181) = 11.09, p = .001, ηp = .06) and, marginally, happiness ( F (1,

2 181) = 3.18, p = .08, ηp = .02). Koreans reported higher levels of affection and marginally higher levels of happiness than Americans when discussing positive experiences. There was no effect of culture on reported experience of gratitude ( F (1, 181) = 0.10, p = .75) or pride ( F (1,

181) = 0.12, p = .73).

Negative discussions. MANOVA results indicate significant overall main effects of

2 condition ( λ = .91, F (8, 358) = 2.17. p = .03, ηp = .05) and culture ( λ = .84, F (4, 179) = 8.76. p

2 < .001, ηp = .16), though there was no significant overall interaction between the two ( λ = .98, F

2 (8, 358) = 0.36. p = .94, ηp = .01). Univariate ANOVA’s indicate the main effect of condition is driven by a significant impact of experimental condition on anger experience ( F (2, 182) = 5.19,

2 p = .006, ηp = .05), with those in the emotion spread condition experiencing more anger than those in the other conditions. Condition did not significantly influence experiences of sadness ( F

(2, 182) = 1.68, p = .19), guilt ( F (2, 182) = 0.98, p = .38), or embarrassment ( F (2, 182) = 0.49, p = .61). The main effect of culture on negative emotions was driven by a significant effect of

2 culture on anger ( F (1, 182) = 29.35, p < .001, ηp = .14) and a marginally significant effect on guilt ( F (1, 182) = 2.95, p = .09). Koreans reporting feeling more anger and marginally less guilt than Americans while discussing negative experiences. Koreans and Americans did not differ in their experiences of sadness ( F (1, 182) = 0.04, p = .83) or embarrassment ( F (1, 182) < 0.01, p =

.99).

Expressive suppression. A 2 (Culture) X 3 (Condition) ANOVA tested the effects of condition and culture on reported expressive suppression. A significant main effect of condition

(F (2, 182) = 3.26, p = .04) indicated participants in the emotion spread condition reported

89 suppressing their emotions more than those in the no spread condition ( p = .05), with those in the control group falling between the two and not differing significantly from either. There was no significant main effect of culture ( F (1, 182) = 2.15, p = .15), however there was a significant interaction between culture and condition ( F (2, 182) = 3.89, p = .02). As shown in Figure 4,

Koreans reported suppressing their emotions more if they were in the emotion spread condition than if they were in the no spread condition ( p = .001) or the control condition ( p = .02).

Meanwhile, condition did not have a significant impact on reported suppression for American participants ( p’s > .10).

Emotional behavior. Analyses were split according to the valence of emotional behaviors because previous studies indicated that the effects of beliefs on reported emotional behaviors were valence-specific. Three 2 X 3 multivariate analyses of covariance (MANOVA’s) tested for effects of culture (Korean, American) and Condition (Emotion spread, No spread,

Control) on (1) positive emotional behavior (i.e., positive emotion word use, nonverbal happiness expressivity) during discussions of positive experiences, (2) positive emotional behavior during discussions of negative experiences, and (3) negative emotional behavior (i.e., negative emotion word use, nonverbal anger expressivity, nonverbal sadness expressivity) during discussions of negative experiences. Negative emotions were expressed at only very low rates during positive discussions and could not be reliably analyzed.

As above, this analytic method was chosen to protect against the chance of Type 1 error.

Ratings of nonverbal behavior did not contain outliers. However, measures of positive and negative emotion word use contained extreme values. To correct for these, extreme values were recoded to the next-highest acceptable value. Testing for correlations among dependent

90 variables, positive emotion word use and nonverbal happiness expressivity during positive discussions were weakly negatively related ( r (184) = -.15, p = .04). During negative discussions, nonverbal expressivity of anger and sadness were moderately correlated with each other ( r (187)

= -.37, p < .001), though neither was significantly associated with negative emotion word use

(anger: r (184) = -.08, p = .30; sadness: r (184) = -.09, p = .24). Finally, positive emotion word use and nonverbal happiness expressivity during negative discussions were not significantly related ( r (184) = .11, p = .13).

Because Koreans and Americans differed in their average age and gender distribution, I ran versions of all main analyses that controlled for these demographic characteristics. Neither factor meaningfully altered the pattern of results in these analyses from that reported below.

Controlling for experienced affection during the analysis of positive discussions and for experienced anger during analysis of negative discussions likewise did not alter the overall patterns of results.

Positive discussions. The results of the MANOVA indicated there was no significant overall main effect of condition ( λ = .99, F (4, 354) = 0.54, p = .71), or culture ( λ = .99, F (2,

177) = 1.38. p = .25), nor was there a significant interaction between culture and condition ( λ =

.99, F (4, 354) = 0.52. p = .72).

Negative discussions. In regards to positive emotional behavior, the results of the

MANOVA indicated there was no significant overall main effect of condition ( λ = .99, F (4, 354)

= 0.53, p = .72), or culture ( λ = .98, F (2, 177) = 1.58. p = .21), nor was there a significant interaction between culture and condition ( λ = .97, F (4, 354) = 1.26. p = .29).

In regards to negative emotional behavior, results of the MANOVA indicated a

2 marginally significant overall main effect of condition ( λ = .94, F (6, 352) = 1.86. p = .086, ηp =

2 .03) and a significant overall effect of culture ( λ = .72, F (3, 176) = 23.26. p < .001, ηp = .28), though there was no significant interaction between the two ( λ = .98, F (6, 352) = 0.46. p = .84).

Univariate analyses of variance (ANOVA) results indicated whether the effects of condition and culture were constrained to particular variables.

Univariate analyses of variance (ANOVA’s) reveal that the main effect of condition was driven by a significant effect of condition on nonverbal anger expressivity ( F (2, 178) = 3.29, p =

2 .04, ηp = .040) and a marginal effect on nonverbal sadness expressivity ( F (2, 178) = 2.39, p =

2 .095, ηp = .03). Post-hoc Bonferroni adjusted mean comparisons reveal that relative to those in the control condition, participants in the emotion spread condition expressed less anger on their faces (p = .04) and marginally less sadness ( p = .10), with participants in the no spread condition falling between the two and not significantly differing from either ( p’s > .10; See Figure 5).

There was no effect of condition on use of negative emotion language ( F (2, 178) = 1.46, p =

.24). The overall effect of culture was driven by its effects on nonverbal expressions of anger ( F

2 2 (1, 178) = 61.84, p < .001, ηp = .26) and sadness ( F (1, 178) = 17.34, p < .001, ηp = .09), such that Koreans nonverbally expressed less anger and sadness than Americans when talking about their negative experiences. There were no cultural differences in negative emotion word use ( F

(2, 178) = 0.29, p = .59).

Expected Support. A 2 (Culture) X 3 (Condition) ANOVA tested the effects of condition and culture on expected support. Results revealed significant main effects of culture ( F

(1, 182) = 27.99, p < .001), with Americans expecting higher quality support overall. A main

92 effect of condition ( F (2, 182) = 3.51, p = .03) indicated participants in the emotion spread condition expected higher quality support than those in the control condition ( p = .04), with those in the no spread condition falling between the two and not significantly differing from either ( p’s

> .10). These main effects were qualified by a significant interaction between the culture and condition ( F (2, 182) = 2.19, p = .04). As shown in Figure 6, both Koreans and Americans expected the highest quality support if they had been led to believe that emotions spread.

However, for Koreans, participants in the emotion spread condition expected significantly better support than those in the no spread condition ( p = .05), with those in the control group falling between the two and not significantly differing from either. For Americans, participants in the emotion spread condition expected significantly better support than those in the control group ( p

= .04). Participants in the no spread condition did not significantly differ from either group, though their mean was closer to that of the emotion spread condition than the control condition. h

Testing the mediating role of interpersonal motives. As reported above, when led to believe emotions spread, participants expressed less anger and sadness and expected better quality support. In addition, Koreans in this condition were more likely to try to suppress their emotions. To test whether interpersonal motives mediated these relationships, I first tested whether the experimental manipulation altered motives to benefit others, strengthen relationships, or receive support among Americans or Koreans using three 2 (culture) X 3

h The inclusion of gender and its interactions with culture and condition reduced the effects of condition and its interaction with culture to non-significance in predicting expected support. However, a significant interaction between gender and condition followed a similar pattern to that of the interaction between culture and gender. While condition had no significant effect on expected support for men, women (like Americans) expected more support if they were in the contagion condition than in the control group, with participants in the no spread condition falling between the two and not significantly differing from either. 93

(condition) ANOVA’s. In this case, a MANOVA was not used because interpersonal motives were more strongly correlated than is recommended (0.34 < r (188) < 0.69).

We found no significant effect of experimental condition on the motives to benefit others

(F (2, 182) = 0.51, p = .60) or strengthen relationships ( F (2, 182) = 0.01, p = .99), but there was a marginally significant effect of condition on the motive to receive support ( F (2, 182) = 2.96, p

= .06), with this motive endorsed most strongly in the control condition. Koreans also endorsed motives to benefit others (F (1, 182) = 5.10, p = .02) and strengthen relationships ( F (1, 182) =

23.59, p < .001) more strongly than Americans, but Koreans and Americans did not differ in their motivations to receive support ( F (1, 182) = 0.86, p = .36). The relationship between culture and the motive to benefit others was qualified by a marginally significant interaction between condition and culture ( F (2, 182) = 2.64, p = .07). Koreans and Americans differed significantly from each other in their endorsement of this motive when in the emotion spread condition ( p =

.002). However, endorsement of this motivation was not significantly impacted by condition for either Koreans or Americans ( p’s > .10). The interaction between culture and condition was non- significant in predicting motives to strengthen relationships ( F (2, 182) = 0.79, p = .45) or receive support (F (2, 182) = 0.71, p = .49).

Only the motivation to receive support was marginally influenced by experimental condition, so I tested whether this motive mediated the effects of experimental condition on use of expressive suppression, nonverbal anger expressivity, nonverbal sadness expressivity, or expected support using a multicategorical mediation model (Hayes & Preacher, 2014; Preacher

& Hayes, 2008). This method used a bootstrapping approach with resampling size of 5000 to obtain estimates of indirect effects of the motive to receive support on the relationship between

94 condition and each outcome. Because the effects of experimental condition on reported suppression and expected support differed for Koreans and Americans, the models testing these outcomes were run separately for each group. The motivation to receive support did not significantly mediate relationships between condition and reported suppression or expected support for either group (95% confidence intervals include zero) or between condition and nonverbal anger expressivity in both groups (95% confidence intervals include zero).

Testing the moderating role of interpersonal motives. This study also tested whether interpersonal motives moderated effects of condition on expressive suppression, nonverbal expressivity, and expected support using hierarchical regression analyses. In the first step, dummy variables corresponding to the emotion spread and no spread conditions were entered along with culture, interactions between conditions and culture, and the interpersonal motives to benefit others, strengthen relationships, and receive support. In step 2, the two-way interactions between interpersonal motives, experimental conditions, and culture were added. Any significant interactions between an interpersonal motive and an experimental condition would indicate a moderating effect of that motive. Step 3 included the three-way interactions between culture, conditions, and motives. A significant three-way interaction would mean that moderating effects of interpersonal motives differ across cultures. Below, I report whether Steps 2 or 3 significantly contribute to the models’ explanatory power and, if so, via which interactions.

Expressive suppression. In the results reported above, Americans’ reports of expressive suppression did not differ across conditions, but Koreans in the emotion spread condition reported trying harder to suppress their emotions than Koreans in the other conditions. The inclusion of interactions between motivations, conditions, and culture in Step 2 did not

95 significantly improve the model’s explanatory power (ΔR 2 = .06), nor did the three-way interactions between these variables ( ΔR 2 < .01), indicating that interpersonal motivations did not significantly moderate the effect of experimental condition on reported suppression.

Nonverbal expressivity. In the results reported above, there was no evidence that experimental condition shifted positive emotional behavior during either discussion. However, when talking about negative experiences, those in the emotion spread condition expressed lower levels of anger and (marginally) lower levels of sadness than those in the control group. Results indicated a small, but statistically significant contribution of the 3-way interactions between motives, condition, and culture to the overall model—with each motive interacting with culture and condition (see Table 11 for all effects of this analysis). These results were complex and differed by motive.

A marginally significant interaction between the motivation to benefit others and emotion spread (β = -0.78, t (170) = -1.91, p = .06) and a significant interaction with culture ( β = -.74, t

(170) = -2.07, p = .04) were qualified by a marginally significant three-way interaction between culture, condition, and the motivation to benefit others (β = 0.86, t (170) = 1.73, p = .09), indicating this motive influenced behavior differently depending upon both culture and condition. For Americans, this motivation was associated with slightly increased nonverbal displays of anger in the emotion spread condition, but was unassociated with behavior in the no spread condition and negatively predicted nonverbal expressivity in the control group. For

Koreans, this motivation was associated with slightly decreased nonverbal anger expressivity in the emotion spread and control conditions, but increased expressivity in the no spread condition

(see Figure 7).

Next, a significant interaction between the motivation to strengthen relationships and emotion spread (β = 1.92, t (170) = 3.07, p = .002) was qualified by a significant interaction between culture, condition, and the motive to strengthen relationships ( β = -2.21, t (170) =

2.79, p = .006). These effects were driven by differing impacts of the motive to strengthen relationships within the emotion spread condition. For Americans led to believe emotions spread, the motive to strengthen relationships predicted increased nonverbal expressions of anger. For

Koreans, this same motivation was unassociated with nonverbal anger expressivity (see Figure

8).

Finally, significant interactions between the motive to receive support and emotion spread (β = -1.49, t (170) = -2.30, p = .02), culture ( β = 1.22, t (170) = 2.17, p = .03) and, marginally, no spread (β = -0.75, t (170) = -1.76, p = .08), were qualified by a significant interaction between culture, condition, and the motivation to receive support ( β = 1.49, t (170) =

1.99, p = .05). As shown in Figure 9, this motivation is unrelated to Koreans’ nonverbal anger displays regardless of experimental condition, while for Americans this motive was associated with slightly increased nonverbal anger expressivity among participants in the control condition, and slightly decreased nonverbal anger expressivity among participants in the emotion spread and no spread conditions (see Figure 9). In sum, interpersonal motives mattered for the nonverbal expression of anger, but their effects were nuanced and depended on both culture and condition.

The addition of interpersonal motives to the regression models predicting nonverbal expressions of sadness did not significantly add explanatory power (ΔR 2 = .07), nor did the inclusion of interactions between motives and conditions or cultures ( ΔR 2 < .01). This indicates

97 that interpersonal motives do not moderate the finding that those led to believe their emotions spread display less sadness.

Expected support. As reported above, Koreans and Americans both expected the highest quality support if they had been led to believe that emotions spread. Including the two-way interactions between motives, condition, and culture contributed significantly to the model’s explanatory power ( ΔR 2 = .08). A significant interaction between culture and the motivation to receive support ( β = 1.08, t (170) = 2.58, p = .01) signified the motivation to receive support was more strongly related to expected support among Koreans than Americans. There were also significant interactions between the motivation to receive support and both experimental conditions (Emotion spread: β = -1.05, t (170) = 2.20, p = .03; No spread: β = -0.90, t (170) =

2.16, p = .03). As shown in Figure 10, the motivation to receive support was unrelated to expectations of support for those in the emotion spread condition. For those in the no spread condition, the more motivated participants were to receive support, the better support they expected. This relationship was even stronger for those in the control condition. Step 3, which introduced three-way interactions between interpersonal motives, conditions, and culture, did not significantly contribute to explanatory power ( ΔR 2 = .01). See Table 12 for all effects of this analysis.

Discussion

The current study was designed to answer three key questions:

1. Are beliefs about the spread of emotions causally related to observable behavior and

emotion regulation?

2. Does belief in the spread of emotions cause people to expect higher quality social

support?

3. Do interpersonal motives explain relationships between beliefs and emotion outcomes?

In regards to the first question, results suggest that beliefs about the spread of emotions do have a causal impact on emotional regulation and on some observable behaviors. Participants led to believe their emotions spread expressed less anger and marginally less sadness when talking about their negative experiences. This was true even though these participants reported feeling higher levels of anger. Koreans led to believe emotions spread also reported increased efforts to suppress their emotions, though Americans’ efforts to suppress did not differ across conditions.

A number of emotional behaviors were unaffected by the experimental manipulation.

Manipulating beliefs about the spread of emotions does not seem to directly influence the positive or negative emotional language people use to talk about their experiences. It is possible that beliefs about the spread of emotion influence whether people choose to discuss their experiences in the first place, but not the emotion language that people use when discussing their experiences. Experimental condition also has no measurable impact on nonverbal displays of happiness. It may be that belief in the spread of happiness does not cause positive emotional behavior, but rather results from frequent positive emotional behavior. This would be consistent with the results of mediation analyses in Study 2, which found that a causal pathway from reported behavior to belief in the spread of happiness fit the data better than the reverse.

Our next question was whether the spread of emotions was perceived as an effective tool for achieving social goals. Toward this end, I examined whether the experimental manipulation influenced the quality of support that participants expected their friend would provide in

99 response to their discussions. Participants led to believe their emotions spread expected higher quality support from their friend, which I take as an indication that the spread of emotions is, at least implicitly, perceived to be an effective route to attaining high quality support. It is interesting to note that for Americans, participants in the emotion spread condition expected significantly better support than did those in the control group, meaning that this manipulation raised expectations beyond those of the cultural “baseline.” It is also notable that expectations of support were quite similar for participants in the emotion spread and no spread conditions, though the latter did not differ significantly from the control group in expectations of support.

The no spread prime did not truly convey that emotions did not spread at all, but rather that emotions spread less than most people believe. This prime may have still subtly cued Americans to think about others while discussing their experiences to a greater extent than they naturally would have. On the other hand, Koreans in the emotion spread condition differed most strongly from those in the no spread condition, with the control group falling between the two. In this case, it seems that being led to believe emotions spread caused Koreans to expect a bit better support than the cultural baseline expected, while being led to believe emotions do not spread caused Koreans to expect slightly worse support. This may be because for Koreans, emotions may be more integrally embedded in social context by default.

The third question of interest was whether interpersonal motives might explain relationships between beliefs about the spread of emotions and outcomes of interest. There was no evidence for a mediating role of interpersonal motives in explaining relationships between beliefs and expressive suppression, nonverbal expressivity, or expected support. In the case of reported suppression, interpersonal motives also did not moderate the effect of the experimental

100 manipulation. Perhaps for Koreans, being led to believe emotions spread causes increased efforts to suppress emotions through more implicit motivations or through the automatic enactment of cultural scripts.

In regards to nonverbal expressivity, results indicate that interpersonal motives influence anger expressivity differently depending upon one’s beliefs and culture. For Americans, relatively other-oriented motivations (to benefit others and strengthen relationships) were associated with more intense and longer-lasting nonverbal displays of anger among participants led to believe emotions spread. In contrast, the relatively self-focused motivation to receive support was associated with decreased nonverbal anger expressivity for Americans in the emotion spread condition. For Americans in the other conditions, these motivations were either more weakly related to behavior or were related to behavior in the opposite direction. For

Koreans, the other-focused motives to benefit others and strengthen relationships had weaker, negative relationships with nonverbal anger expression among those in the emotion spread condition, but these motivations were associated with increased nonverbal anger expressivity among Koreans in the no spread condition. The relatively self-focused motivation to receive support was largely unrelated to nonverbal anger expressivity in each condition.

It is possible these differences reflect culturally shaped ideas about how to accomplish these goals. For Koreans, the relationships between motivations and behavior make sense given the norms for social behavior that are encouraged in relatively interdependent cultural contexts.

When emotions are perceived to have a strong impact on others, expressing one’s anger may seem to conflict with salient interpersonal goals to benefit or protect others and to feel close and connected to others. When Koreans believe their emotions will have less of an impact on others,

101 the expression of anger may seem less destructive toward these goals. Meanwhile, the relatively self-focused goal does not seem to influence behavior strongly in this context. For Americans, the opposite may be true. It may seem counterintuitive that spreading one’s anger might be considered conducive to making others feel good or strengthening relationships, but these relationships make more sense given American cultural norms which place value on the open expression of emotion in close relationships. In contrast, spreading one’s anger may be believed to hinder one’s ability to receive support. However, it is also possible that individuals express their emotions less when they believe their emotions spread because they believe they will receive high quality support regardless of their expressivity.

In regards to expectations of support, analysis of interpersonal motives indicated a moderating effect of the motivation to receive support. Interestingly, being led to believe emotions spread seemed to reduce the relevance of this motivation in driving expectations of support. For participants in the control condition and for those led to believe emotions do not spread, motivations to receive support were associated with expectations of better quality support. However, for those led to believe emotions spread, the motivation to receive support was unrelated to expectations. It may be that the spread of emotions is perceived to elicit support fairly automatically, leaving little need for motivated effort.

Some interesting patterns of cultural differences also emerged in this study. As described in the introduction, a relatively interdependent view of the self tends to be encouraged in Korean context, while a relatively independent view of the self is encouraged in American cultural context. In this sample, Americans did endorse an independent self-construal more strongly than

Koreans, but Americans also endorsed an interdependent self-construal more strongly than

102

Koreans. However, the measure of interdependence was also not reliable in the current sample of

Korean participants. Explicit measures of individualism and collectivism have been shown to have low reliability in Korean samples in past research, as well (Oyserman et al., 2002); however, the low reliability of this measure makes it difficult to interpret relationships (or lack thereof) between this construct and other variables of interest. In line with past research, Koreans reported experiencing higher levels of the socially engaging emotion of affection while talking about their positive experiences (Eid & Diener, 2001; Kitayama, et al., 2000; Kitayama et al.,

2006) and endorsed the motivations to make others feel good and strengthen relationships more strongly than Americans (Taylor et al., 2004). Surprisingly, however, Koreans also reported feeling anger more intensely than Americans when discussing negative experiences, a pattern that counters research finding Japanese tend to experience anger less frequently and less intensely than Americans (Boiger et al., 2014; Boiger et al., 2013). Despite this, Koreans still expressed their anger and sadness less intensely than Americans. The suppression of anger expression is a particularly strong Korean script—as reflected in the existence of a cultural syndrome called Hwa Byung, a disorder whose primary symptoms are somatic presentations of anger (Choi, Chentsova-Dutton, & Parrott, 2016). The strong difference between Koreans’ experience and expression of anger in this study may reflect a cultural script encouraging the suppression of emotional anger displays.

This study contained a number of important limitations. Perhaps most notably, although this study is motivated by a perceived need to situate research on beliefs about the consequences of emotions in social context, this experiment is not a true interpersonal interaction. Past research has shown that imagined and real social presence can influence behavior in similar ways (e.g.,

103

Jakobs et al., 1999). Furthermore, controlling for any effects the interaction partner may have had on behavior may have allowed a clearer signal from the experimental manipulation. Still, much of the complexity of true interactions was lost in the current study. This research is also limited by its sample. Not only is the sample fairly small, but the use of college student samples invites questions about the generalizability of results outside of this population. These and other limitations of the current work will be discussed in more depth in the general discussion.

104

Table 11. Hierarchical regressions predicting nonverbal anger expressivity

B SE β t ΔR 2 Step 1 .31*** Step 2 .03 Step 3 .03* (Constant) -2.58 1.32 -1.95* ** Culture (Ref: Korean) -1.67 1.65 -0.42 -1.01 *** Emotion spread -1.42 2.34 -0.33 -0.61 No spread -0.48 1.55 -0.11 -0.31 Emotion spread * Korean -1.64 3.19 -0.30 -0.51 No spread * Korean -0.05 0.68 -0.01 -0.07 Benefit Others -0.50 0.37 -0.25 -1.34 Strengthen Relationships -0.51 0.41 -0.27 -1.24 Receive Support -0.31 0.32 -0.14 -0.97 Emotion spread * Benefit Others -1.16 0.61 -0.78 -1.91 t No spread * Benefit Others -0.21 0.45 -0.15 -0.47 Emotion spread * Strengthen -2.03 0.66 -1.92 -3.07*** Relationships No spread * Strengthen Relationships -0.72 0.51 -0.68 -1.40 Emotion spread * Receive Support -1.41 0.61 -1.49 -2.30* No spread * Receive Support -0.72 0.41 -0.75 -1.76 t Culture * Benefit Others -0.92 0.44 -0.74 -2.07* Culture * Strengthen Relationships -0.46 0.40 -0.55 -1.15 Culture * Receive Support -1.09 0.50 -1.22 -2.17* Culture * Emotion spread * Benefit -1.47 0.85 -0.86 -1.73 t Others Culture * Emotion spread * Strengthen -2.72 0.98 -2.21 -2.79** Relationships Culture * Emotion spread * Receive -1.74 0.88 -1.49 -1.99* Support t p < .10; * p < .05; ** p < .01; *** p < .001

105

Table 12. Hierarchical regressions predicting expected support

B SE β t ΔR 2 Step 1 .30*** Step 2 .08* (Constant) -3.69 0.57 -6.49*** Culture (Ref: Korean) -1.91 0.64 -1.06 -2.96*** Emotion spread -3.02 0.77 -1.55 -3.92*** No spread -1.59 0.69 -0.82 -2.31* Emotion spread * Korean -0.09 0.30 -0.04 -0.28 No spread * Korean -0.47 0.30 -0.19 -1.56 Benefit Others -0.06 0.15 -0.07 -0.41 Strengthen Relationships -0.01 0.17 -0.01 -0.06 Receive Support -0.43 0.14 -0.44 -3.14*** Emotion spread * Benefit Others -0.07 0.20 -0.10 -0.34 No spread * Benefit Others -0.06 0.20 -0.09 -0.30 Emotion spread * Strengthen -0.18 0.23 -0.38 -0.80 Relationships No spread * Strengthen Relationships -0.13 0.23 -0.27 -0.56 Emotion spread * Receive Support -0.45 0.20 -1.05 -2.20* No spread * Receive Support -0.39 0.18 -0.90 -2.16* Culture * Benefit Others -0.04 0.16 -0.06 -0.21 Culture * Strengthen Relationships -0.41 0.16 -1.08 -2.58** Culture * Receive Support -0.08 0.19 -0.21 -0.45 Step 3 .01 t p < .10; * p < .05; ** p < .01; *** p < .001

106

2 Reported Suppression Reported

1 Korean American

Emotion Spread No Spread Control

Figure 4: Koreans’ and Americans’ mean levels of reported emotion regulation effort in each experimental condition. Error bars represent standard errors.

107

1 Nonverbal Expressivity Nonverbal

0 Anger Expressivity Sadness Expressivity Emotion Spread No Spread Control

Figure 5: Mean levels of nonverbal expressivity in each condition while discussing negative emotional experiences. Error bars represent standard errors.

108

Expected Support Expected 5

4 Korean American

Emotion Spread No Spread Control

Figure 6: Koreans’ and Americans’ mean levels of expected support in each experimental condition. Error bars represent standard errors.

109

1 Nonverbal Anger Expressivity Anger Nonverbal 0 1 2 3 4 5 6 7 Motivation to Benefit Others

American Emotion Spread American No Spread American Control Korean Emotion Spread Korean No Spread Korean Control

Figure 7. Relationships between the motivation to make others feel good and nonverbal anger expressivity in each condition, separated by culture.

110

5 4

2 1

Nonverbal Anger Expressivity Anger Nonverbal 0 1 2 3 4 5 6 7 Motivation to Strengthen Relationships

American Emotion Spread American No Spread American Control

Korean Emotion Spread Korean No Spread Korean Control

Figure 8. Relationships between the motivation to strengthen relationships and nonverbal anger expressivity in each condition, separated by culture.

111

6 5 4 3 2 1 0 Nonverbal Anger Expressivity Anger Nonverbal 1 2 3 4 5 6 7 Motivation to Receive Support American Emotion Spread American No Spread American Control

Korean Emotion Spread Korean No Spread Korean Control

Figure 9. Relationships between the motivation to receive support and nonverbal anger expressivity in each condition, separated by culture.

112

3 Expected Support Expected 2

1 1 2 3 4 5 6 7 Motivation to Receive Support

Emotion Spread No Spread Control

Figure 10. Relationships between the motivation to receive support and expectations of support in each condition.

113

Chapter 5: General Discussion

The research presented in this dissertation introduces a novel construct to the literature on beliefs about emotions: beliefs about the interpersonal spread of emotions. Research investigating beliefs about the interpersonal consequences of emotions is underdeveloped, despite widespread agreement that emotions are fundamentally social in nature and that beliefs about emotions are important. This research is meant to address this gap in the existing psychological literature. Across three studies, I build an understanding of how people conceptualize the spread of their emotions to others, how these beliefs relate to perceived and observable emotional behavior, and how interpersonal goals contribute to these relationships.

First, this research established some similarities in how individuals across cultures conceptualize the spread of emotions. Beliefs about the spread of happiness and sadness could be meaningfully separated in both American and Korean samples. Furthermore, both Americans and Koreans were more likely to endorse the idea that their happiness spread than that their sadness spread, though the difference between these was narrower among Koreans. This research also uncovered some differences in conceptions of these beliefs across cultures. For Americans, beliefs about the spread of emotions are largely valence specific—people that believe their happiness spread also tend to believe their excitement and their calm feelings spread, while those believing their sadness spreads likewise believe their anger will spread. For Koreans, the distinction based on valence was less clear-cut, with beliefs about the spread of happiness being distinct from beliefs about the spread of most other emotions.

Next, this research found that beliefs about the spread of emotions were related to emotion regulation and behavior in all three studies, though the direction of these relationships

114 was not consistent across all studies. In Studies 1-2, beliefs about the spread of emotions were associated with increased reports of one’s own emotional regulation and expressivity. Those who believed their happiness spread reported suppressing their emotions less, being more nonverbally expressive of their positive emotions, and talking about their positive experiences more. Similar patterns were found for beliefs about the spread of sadness—those who believed their sadness spread reported suppressing their emotions less, being more nonverbally expressive of their negative emotions, and disclosing their negative experiences to others more often. However, when beliefs about the spread of emotions were manipulated and observable behavior assessed in

Study 3, being led to believe that emotions spread was associated with increased rather than decreased reported use of emotion regulation (though only in Koreans), as well as with corresponding decreases in the nonverbal expression of anger and sadness. Contrary to the expectation based on people’s reports of their behavior, the manipulation of this belief had no significant impacts on positive nonverbal behavior, nor did it influence the emotion language that people use to talk about their positive or negative experiences. What factors account for these differences in observed patterns of results across studies?

It may be that when beliefs about the spread of emotions are measured via self-report, participants are thinking about the spread of emotions in more abstract terms. The experimental manipulations, however, force participants to think through the actual consequences of spreading emotions. When people are forced to confront the consequences of spreading their negative emotions to a specific close other at a more conscious level, they may be more strongly motivated to regulate their emotions, particularly their negative emotions, and to express less negative emotion. Since the other person was not physically present during this study, this study

115 may have been a fairly conservative test of this effect—had a true interaction occurred, any concerns, implicit or explicit, linked to the spread of emotions would likely have had a stronger influence on behavior.

An alternative explanation for these diverging patterns is that beliefs about the spread of emotions co-occur with increased perceptions of emotional behavior, but do not cause them.

When beliefs about the spread of emotions are actively altered, their relationships to emotional behavior and regulation may therefore differ from those that exist when beliefs are measured as they naturally occur. Indeed, one may come to believe emotions spread as a result of expressing emotions often and seeing the process of emotional contagion at work. In support of this explanation, results of mediation analyses in Study 2 indicated that models predicting belief in the spread of happiness from emotional behavior fit the data well. Perhaps beliefs about the spread of emotions, especially happiness, are more naturally the result rather than the cause of emotional behavior.

Finally, in addressing the discrepancy between reported sharing of positive and negative experiences and actual emotion language used in discussing experiences, it should be noted that these measures are fundamentally distinct. In Studies 1-2, beliefs about the spread of emotions were associated with the tendencies to discuss positive and negative experiences with others, but participants were not asked to report on the emotional language that they use in these discussions. In Study 3, all participants were directed to discuss their experiences. It is possible that beliefs about the spread of emotion are causally related to the rates at which people discuss their emotional experiences and seek support from others, but not with the emotion language that they use once they choose to do so.

116

This research also tested the role of interpersonal goals in explaining relationships between beliefs and behavior. In particular, I tested the effects of the goal to make others feel good (which is primarily focused on others), to strengthen relationships (which may benefit both the self and others), and to receive support (which is primarily focused on the self). I expected that relatively other-focused goals would be more closely related to emotional behavior among

Koreans and relatively self-focused goals would be more closely related to emotional behavior among Americans. In Study 2, interpersonal motives mediated relationships between beliefs about the spread of emotions and reported behavior similarly in both the U.S. and Korea. People who believed their happiness spread reported expressing positive emotions more due to an increased motivation to strengthen relationships with others. People who believed their sadness spread reported expressing negative emotions more due to an increased motivation to receive support from others.

Based on the results of Studies 1-2, I believed people perceived the spread of emotions as a useful means of achieving their goals to build relationships and receive support. In Study 3, those led to believe their emotions spread expected their friends to provide them with higher quality support than did their peers in the other conditions. This suggests that people perceive the spread of emotions to be useful in some interpersonal contexts. When testing the role of interpersonal motives in explaining this effect of experimental condition, beliefs about the spread of emotions altered the relevance of the motive to receive support to people’s expectations of support. For people in the control condition or those led to believe emotions do not spread easily, being motivated to receive support is associated with expecting support. When people are motivated to get high quality support, they believe they will get it. However, the motivation to

117 receive support was unrelated to expectations of support when people were led to believe their emotions spread. It seems that when participants are led to believe their emotions spread, they believe they will receive high-quality support from their friends whether they want it or not.

Unfortunately, expectations of relational closeness were not measured in Study 3, so I could not explicitly test whether beliefs about the spread of emotions are perceived to facilitate achievement of the goal to strengthen relationships

This research was conducted in South Korea because there were theoretical reasons to expect beliefs about the spread of emotions to relate to emotional behavior differently in this context. Because relatively interdependent views of the self tend to prevail in East Asian contexts like Korea, Koreans might endorse other-focused motivations more strongly than

Americans. These other-focused motives may lead Koreans to regulate their emotions more and express their emotions less when they believed emotions spread. Measures of self-construal revealed that Koreans did not endorse an interdependent self-construal more highly than

Americans in either study. Still, Koreans did endorse the relatively other-focused motivation to make others feel good more strongly than Americans in Studies 2-3, and the motivation to strengthen relationships more strongly in Study 3.

Despite differences in the strength of relatively other-focused motivations, Study 2 found little evidence of cultural differences in the relationships between beliefs about the spread of emotions and reported behavior. In this study, culture did not moderate any of the relationships observed between beliefs about the spread of emotions and reported emotional behaviors. I found some support for the hypothesis in Study 3. The effects of experimental condition on emotional behavior were consistent across cultures for most dependent variables in this study. In the single,

118 but notable, exception, Koreans led to believe their emotions spread reported trying harder to regulate their emotions than did Koreans in the other two conditions. Americans’ regulatory efforts were unaffected by this manipulation. However, this effect could not be explained by a stronger endorsement of other-focused interpersonal motivations, at least not those measured in the current research. It may be that for Koreans, being forced to think through the actual consequences of spreading drives attempts to regulate behavior in more nonconscious ways.

In addition to this effect of culture on regulatory effort, interpersonal motivations resulted in different expressions of anger for Koreans and Americans, depending on their beliefs about the spread of emotions. Relatively other-focused motives to benefit others and strengthen relationships are associated with increased expressivity for Americans, but decreased expressivity for Koreans. This effect may reflect differing norms for the expression of anger—in

Korea, expressing one’s anger may be perceived as especially counterproductive if it is likely to spread and one’s goals are to bond with others and protect their feelings. For Americans, the opposite may be the case—expressing one’s anger may be perceived as useful if it is likely to spread and one’s goals are to bond with others and make them feel good. The relatively self- focused motive to receive support predicted decreased expressions of anger among Americans led to believe emotions spread, but was unrelated to behavior among Koreans in all conditions.

Similar to how the belief that emotions spread reduced the relevance of this motivation in predicting expected support, perhaps the belief that emotions spread reduces Americans’ perceptions of the need for open expression in pursuing support.

Cultural differences related to other variables were sometimes counterintuitive. In Study

2, Americans and Koreans did not differ in most domains of reported emotional behavior.

119

However, Koreans reported seeking instrumental support more often than Americans.

Furthermore, while Koreans did endorse a motivation to make others feel good more strongly than Americans, as expected, they also endorsed the relatively self-focused motivation to receive support more strongly than Americans in Study 2. In Study 3, Koreans expressed less anger and sadness than Americans when talking about their negative experience, as expected. However,

Koreans also reported experiencing more anger than Americans did during these same discussions.

Overall, Koreans and Americans were more similar to each other than they were different. This was especially true in Study 2, which relied on self-reports. In study 3, however, the cultural differences that did emerge were largely in directions consistent with predictions based on an interdependent view of the self, even though self-reported self-construal did not reflect this orientation. Relative to Americans, Koreans were less nonverbally expressive of their anger and endorsed other-focused motivations more strongly. When led to believe emotions spread, they tried to suppress their emotions more and expressed their anger less if they were also highly motivated to benefit others or strengthen relationships. As discussed in Chapter 3, it is possible that certain characteristics of the Korean sample (i.e., urban, young, college students) or American sample (i.e., students at a Jesuit university) may have reduced the ability to detect cultural differences that are present in broader society. However, since culture is always changing, it is also possible that counterintuitive cultural differences in these data accurately reflect modern beliefs, motivations, and expressive norms of these two groups of college students.

120

Limitations and Future Directions

There are many avenues for building upon and improving the research described in this dissertation. One limitation of the current research lies in the restricted range of emotions results can be generalized to. Some preliminary findings suggesting that results may generalize from beliefs about the spread of sadness to beliefs about the spread of anger (Studies 2-3). Still, more work should be done to test people’s beliefs about how other emotions impact others, especially other emotions known to have different meanings across cultures (i.e., pride, shame). It would also be interesting to sample a wider range of potential consequences of one’s emotions for others. This research examined beliefs about whether one’s happiness makes others happy, and one’s sadness makes others sad, but one can imagine situations where emotions may elicit complementary, or even contradictory, emotions in others. For example, I would like to test people’s beliefs about how their own pride or shame makes others feel, and how these beliefs influence their tendencies to engage in behaviors like boasting or apologizing.

An additional area of concern in the current research lies in the generalizability of the sample. Psychological research has long been criticized for its reliance on student samples and

WEIRD populations (Henrich et al., 2010), and the current research is subject to these same criticisms. Though Study 1 included an adult sample from Amazon mechanical Turk and Studies

2-3 included samples collected in South Korea, these groups may not be representative of the populations of interest. In my post-doctoral work, I will be working primarily with Detroit’s

African American population, and I am interested in examining these beliefs in this unique community setting. I hope to begin this work with a qualitative investigation, asking people how they believe their positive and negative emotions impact the people around them. This way, I can

121 see whether the belief that emotions spread is a salient one in this context, or whether other beliefs are more likely drivers of emotional behavior in interpersonal context.

I have argued that studying beliefs about the spread of emotions is important because a nuanced understanding of how people conceptualize the interpersonal consequences of their emotions is missing from the existing psychological literature. Emotions unfold in the context of interactions and relationships, and people’s beliefs about the consequences of their emotions for others and for their relationships may contribute to emotional behavior in social context. Though this research is grounded in the importance of social context in shaping emotions, I examined only one side of the interaction in the current research. One avenue for future research lies in examining beliefs about the spread of emotions from the other side of the interaction: those catching emotions. In particular, I am interested to know whether people who believe they will catch emotions perceive themselves to be better support providers, as people who believe they will spread emotions expect to receive higher quality support.

In a similar vein, examining these beliefs in the context of dynamic interactions and specific relationship contexts is an important next step. This is a key limitation of the current research, as qualities of interaction partners and relationships undoubtedly play important roles in shaping beliefs, motives, behavior, and the relationships between them. In Study 2, data suggested people were primarily thinking about their friends and family when considering whether their emotions spread, and Study 3 explicitly asked participants to think about a close friend. Salient interpersonal goals differ depending upon who one is interacting with. Beliefs about the spread of emotion may relate to emotional behavior differently when people interact in workplaces, healthcare settings, or schools than when they interact with close friends. Studies

122 that systematically manipulate characteristics of relationships and interaction partners are an important next step in this research.

One area in this domain I am particularly interested in exploring is how these constructs relate to each other when one’s interaction partner is perceived to be vulnerable or unstable in some way. When thinking about close friends in general, or instructed to select a friend to share emotional experiences with, most participants probably imagine others they presume to be emotionally healthy and resilient. Perhaps motivations to benefit others (or protect them) is only associated with increased expressivity when people can assume their interaction partners will catch emotions to only a moderate degree or for only a short amount of time. However, if interaction partners are perceived to be somehow psychologically vulnerable, through physical or mental illness, for example, even Americans may be more motivated to regulate their expression of negative emotions if they believe emotions spread. In my post-doctoral work, I intend to examine these and other beliefs about the interpersonal consequences of emotions among the friends, family, and physicians of people with cancer diagnoses.

Studying beliefs about the spread of emotions in the context of dynamic interpersonal interactions would also allow for an examination of whether belief in the spread of emotions benefits social interactions. Sharing positive emotions can facilitate bonding and affiliation, suggesting that this belief and its relationship to increased positive expressivity may indeed be an adaptive one (Collins & Miller, 1994; Glover & Parry, 2008; McPherson et al., 2001). Similarly, expressing negative emotions can elicit empathy and support (Cobb, 1976; Thoits, 2011).

However, the receipt of social support can backfire, leaving support seekers feeling worse than they did before the interaction (Taylor, 2011). Examining the influence of beliefs about the

123 spread of emotions on behavior during interpersonal interactions or in the context of long-term relationships would provide insight into whether the relationships observed in this research between beliefs, motives, and behavior are beneficial.

Identifying predictors of healthy emotional engagement is of special significance in the context of coping with disease—when patients, physicians, family, and friends are tasked with navigating deeply emotional experiences, and doing so effectively can mean improved health and well-being for patients. In particular, beliefs about the spread of emotions are a promising set of constructs in understanding patient-physician interactions, where they might inform how physicians engage emotionally with their patients. For example, if a physician believes herself capable of spreading positive feelings to a patient, and believes positive feelings will be beneficial to the patient, she should be more likely to express positivity in their interaction.

Beliefs are a promising set of constructs in this regard—beliefs lend themselves to change, making them a potentially useful lever in interventions promoting effective social interactions and maximizing well-being. At the Karmanos Cancer Institute, I hope to probe physicians’ beliefs about their own and their patients’ emotions and explore how these beliefs might influence patient-physician interactions and patients’ affective responses to their cancer diagnoses and treatments.

Conclusions

Despite its limitations, the current research contributes to our understanding of the important role that beliefs about the social consequences of emotions—particularly beliefs about the spread of emotions—play in guiding emotional behavior. In particular, this work suggests that understanding what people believe about the spread of their emotions to others can be useful

124 in predicting how they perceive their own emotional behavior, how they attempt to manage their emotions, and how they actually express their emotions. Furthermore, this research demonstrates that beliefs about the spread of emotions can be altered, and subsequently impact important interpersonal and affective phenomena, including expectations of others and observable emotional behavior. What people think about the social consequences of emotions matters, and scientific understanding of emotion benefits by incorporating experience, expression, and cognitive representations of emotions.

125

Appendix: Experimental Manipulations

126

127

128

129

130

131

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders

(5th ed.). Arlington, VA: American Psychiatric Publishing.

Anderson, C., Keltner, D., & John, O. P. (2003). Emotional convergence between people over

time. Journal of Personality and Social Psychology , 84 , 1054–1068.

http://doi.org/10.1037/0022-3514.84.5.1054

Bastian, B., Koval, P., Erbas, Y., Houben, M., Pe, M., & Kuppens, P. (2015). Sad and alone:

Social expectancies for experiencing negative emotions are linked to feelings of

loneliness. Social Psychological and Personality Science , 6, 146–178.

http://doi.org/10.1177/1948550614568682

Bastian, B., Kuppens, P., De Roover, K., & Diener, E. (2014). Is valuing positive emotion

associated with life satisfaction? Emotion , 14 , 639–645. http://doi.org/10.1037/a0036466

Bastian, B., Kuppens, P., Hornsey, M. J., Park, J., Koval, P., & Uchida, Y. (2012). Feeling bad

about being sad: The role of social expectancies in amplifying negative mood. Emotion ,

12 , 69–80. http://doi.org/10.1037/a0024755

Batson, C. D. (1991). The Altruism Question: Toward a Social-Psychological Answer . Hillsdale,

NJ: Lawrence Erlbaum Associates, Inc.

Batson, C. D., C., Duncan, B. D., Ackerman, P., Buckley, T., & Birch, K. (1981). Is empathic

emotion a source of altruistic motivation? Journal of Personality and Social Psychology ,

40 , 290–302. http://doi.org/10.1037/0022-3514.40.2.290

132

Boiger, M., Güngör, D., Karasawa, M., & Mesquita, B. (2014). Defending honour, keeping face:

Interpersonal affordances of anger and shame in Turkey and Japan. Cognition and

Emotion , 28 , 1255–1269. http://doi.org/10.1080/02699931.2014.881324

Boiger, M., & Mesquita, B. (2012). The construction of emotion in interactions, relationships,

and cultures. Emotion Review , 4, 221–229. http://doi.org/10.1177/1754073912439765

Boiger, M., Mesquita, B., Tsai, A. Y., & Markus, H. (2012). Influencing and adjusting in daily

emotional situations: A comparison of European and Asian American action styles.

Cognition and Emotion , 26, 332–340. http://doi.org/10.1080/02699931.2011.572422

Boiger, M., Mesquita, B., Uchida, Y., & Barrett, L. F. (2013). Condoned or condemned: The

situational affordance of anger and shame in the United States and Japan. Personality and

Social Psychology Bulletin , 39 , 540–553. http://doi.org/10.1177/0146167213478201

Bourg Carter, S. (2012, October 12). Emotions Are Contagious—Choose Your Company

Wisely. Retrieved April 23, 2016, from http://www.psychologytoday.com/blog/high-

octane-women/201210/emotions-are-contagious-choose-your-company-wisely

Butler, E. A., Lee, T. L., & Gross, J. J. (2007). Emotion regulation and culture: Are the social

consequences of emotion suppression culture-specific? Emotion , 7, 30–48.

http://doi.org/10.1037/1528-3542.7.1.30

Campbell-Sills, L., Barlow, D. H., Brown, T. A., & Hofmann, S. G. (2006). Acceptability and

suppression of negative emotion in anxiety and mood disorders. Emotion , 6, 587–595.

http://doi.org/10.1037/1528-3542.6.4.587

133

Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: A

theoretically based approach. Journal of Personality and Social Psychology , 56 , 267–

283. http://doi.org/10.1037/0022-3514.56.2.267

Chentsova-Dutton, Y. E., Senft, N., & Ryder, A. G. (2013). Listening to negative emotions: How

culture constrains what we hear. In Parrott, W. G. (Ed.), The Positive Side of Negative

Emotions (pp. 146–178). New York, NY, US: Guilford Press.

Chentsova-Dutton, Y. E., & Tsai, J. L. (2010). Self-focused attention and emotional reactivity:

The role of culture. Journal of Personality and Social Psychology , 98 , 507–519.

http://doi.org/10.1037/a0018534

Choi, E., Chentsova-Dutton, Y., & Parrott, W. G. (2016). The effectiveness of somatization in

communicating distress in Korean and American cultural contexts. Frontiers in

Psychology , 7. http://doi.org/10.3389/fpsyg.2016.00383

Chow, P. I., Berenbaum, H., & Flores, L. E. (2013). The role of perceived utility of emotion in

interpersonal sensitivity and depression. Journal of Social and Clinical Psychology , 32 ,

859–877. http://doi.org/10.1521/jscp.2013.32.8.859

Choy, Y., Schneier, F. R., Heimberg, R. G., Oh, K.-S., & Liebowitz, M. R. (2008). Features of

the offensive subtype of Taijin-Kyofu-Sho in U.S. and Korean patients with DSM-IV

social anxiety disorder. Depression and Anxiety , 25 , 230–240.

http://doi.org/10.1002/da.20295

Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine , 38 , 300–

314.

134

Cole, P. M., Tamang, B. L., & Shrestha, S. (2006). Cultural variations in the socialization of

young children’s anger and shame. Child Development , 77 , 1237–1251.

http://doi.org/10.1111/j.1467-8624.2006.00931.x

Collins, N. L., & Miller, L. C. (1994). Self-disclosure and liking: A meta-analytic review.

Psychological Bulletin , 116 , 457–475. http://doi.org/10.1037/0033-2909.116.3.457

Dakof, G. A., & Taylor, S. E. (1990). Victims’ perceptions of social support: What is helpful

from whom? Journal of Personality and Social Psychology , 58 , 80–89.

http://doi.org/10.1037/0022-3514.58.1.80

De Castella, K., Goldin, P., Jazaieri, H., Ziv, M., Heimberg, R. G., & Gross, J. J. (2014).

Emotion beliefs in social anxiety disorder: Associations with stress, anxiety, and well-

being. Australian Journal of Psychology , 66 , 139–148. http://doi.org/10.1111/ajpy.12053

Diener, E., Suh, E. M., Smith, H., & Shao, L. (1995). National differences in reported subjective

well-being: Why do they occur? Social Indicators Research , 34 , 7–32.

http://doi.org/10.1007/BF01078966

Doherty, R. W. (1997). The Emotional Contagion Scale: A measure of individual differences.

Journal of Nonverbal Behavior , 21 , 131–154. http://doi.org/10.1023/A:1024956003661

Dzokoto, V., Senft, N., Kpobi, L., & Washington-Nortey, P.-M. (2016). Their hands have lost

their bones: Exploring cultural scripts in two West African affect lexica. Journal of

Psycholinguistic Research , 1–25. http://doi.org/10.1007/s10936-016-9415-5

Ehring, T., Tuschen-Caffier, B., Schnülle, J., Fischer, S., & Gross, J. J. (2010). Emotion

regulation and vulnerability to depression: Spontaneous versus instructed use of emotion

suppression and reappraisal. Emotion , 10 , 563–572. http://doi.org/10.1037/a0019010

135

Eid, M., & Diener, E. (2001). Norms for experiencing emotions in different cultures: Inter- and

intranational differences. Journal of Personality and Social Psychology , 81 , 869–885.

http://doi.org/10.1037/0022-3514.81.5.869

Eifert, G. H., Zvolensky, M. J., Sorrell, J. T., Hopko, D. R., & W, C. (1999). Predictors of self-

reported anxiety and panic symptoms: An evaluation of anxiety sensitivity, suffocation

fear, heart-focused anxiety, and breath-holding duration. Journal of Psychopathology and

Behavioral Assessment , 21 , 293–305. http://doi.org/10.1023/A:1022116731279

Eke, M., & McNally, R. J. (1996). Anxiety sensitivity, suffocation fear, trait anxiety, and breath-

holding duration as predictors of response to carbon dioxide challenge. Behaviour

Research and Therapy , 34 , 603–607. http://doi.org/10.1016/0005-7967(96)00044-7

Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use

of exploratory factor analysis in psychological research. Psychological Methods , 4, 272–

299. http://doi.org/10.1037/1082-989X.4.3.272

Ford, B. Q., Dmitrieva, J. O., Heller, D., Chentsova-Dutton, Y., Grossmann, I., Tamir, M., …

Mauss, I. B. (2015). Culture shapes whether the pursuit of happiness predicts higher or

lower well-being. Journal of Experimental Psychology: General , 144 , 1053–1062.

http://doi.org/10.1037/xge0000108

Ford, B. Q., Mauss, I. B., & Gruber, J. (2015). Valuing happiness is associated with bipolar

disorder. Emotion , 15 , 211–222. http://doi.org/10.1037/emo0000048

Ford, B. Q., Shallcross, A. J., Mauss, I. B., Floerke, V. A., & Gruber, J. (2014). Desperately

seeking happiness: Valuing happiness is associated with symptoms and diagnosis of

136

depression. Journal of Social and Clinical Psychology , 33 , 890–905.

http://doi.org/10.1521/jscp.2014.33.10.890

Fredrickson, B. L. (1998). What good are positive emotions? Review of General Psychology , 2,

300–319. http://doi.org/10.1037/1089-2680.2.3.300

Fredrickson, B. L. (2004). The broaden–and–build theory of positive emotions. Philosophical

Transactions of the Royal Society of London B: Biological Sciences , 359 , 1367–1377.

http://doi.org/10.1098/rstb.2004.1512

Fung, H. (1999). Becoming a moral child: The socialization of shame among young Chinese

children. Ethos , 27 , 180–209. http://doi.org/10.1525/eth.1999.27.2.180

Gable, S. L., & Reis, H. T. (2010). Good news! Capitalizing on positive events in an

interpersonal context. Advances in Experimental Social Psychology , 42 , 195–257.

Glover, T. D., & Parry, D. C. (2008). Friendships developed subsequent to a stressful life event:

The interplay of leisure, social capital, and health. Journal of Leisure Research , 40 , 208–

230.

Goetz, J. L., Keltner, D., & Simon-Thomas, E. (2010). Compassion: An evolutionary analysis

and empirical review. Psychological Bulletin , 136 , 351–374.

http://doi.org/10.1037/a0018807

Gosling, S. D., Rentfrow, P. J., & Swann, W. B. (2003). A very brief measure of the Big-Five

personality domains. Journal of Research in Personality , 37 , 504–528.

http://doi.org/10.1016/S0092-6566(03)00046-1

137

Graham, S. M., Huang, J. Y., Clark, M. S., & Helgeson, V. S. (2008). The positives of negative

emotions: Willingness to express negative emotions promotes relationships. Personality

and Social Psychology Bulletin , 34 , 394–406. http://doi.org/10.1177/0146167207311281

Gross, J. J. (2002). Emotion regulation: Affective, cognitive, and social consequences.

Psychophysiology , 39 , 281–291. http://doi.org/10.1017/S0048577201393198

Gross, J. J., & John, O. P. (1995). Facets of emotional expressivity: Three self-report factors and

their correlates. Personality and Individual Differences , 19 , 555–568.

http://doi.org/10.1016/0191-8869(95)00055-B

Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes:

Implications for affect, relationships, and well-being. Journal of Personality and Social

Psychology , 85 , 348–362. http://doi.org/10.1037/0022-3514.85.2.348

Grossmann, I., Huynh, A. C., & Ellsworth, P. C. (2015). Emotional complexity: Clarifying

definitions and cultural correlates. Journal of Personality and Social Psychology .

http://doi.org/10.1037/pspp0000084

Haga, S. M., Kraft, P., & Corby, E.-K. (2009). Emotion regulation: Antecedents and well-being

outcomes of cognitive reappraisal and expressive suppression in cross-cultural samples.

Journal of Happiness Studies , 10 , 271–291.

Han, J. Y., Shah, D. V., Kim, E., Namkoong, K., Lee, S.-Y., Moon, T. J., … Gustafson, D. H.

(2011). Empathic exchanges in online cancer support groups: Distinguishing message

expression and reception effects. Health Communication , 26 , 185–197.

http://doi.org/10.1080/10410236.2010.544283

138

Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1994). Emotional Contagion . New York, NY,

US: Cambridge University Press.

Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical

independent variable. British Journal of Mathematical and Statistical Psychology , 67 ,

451–470. http://doi.org/10.1111/bmsp.12028

Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a universal need for

positive self-regard? Psychological Review , 106 , 766–794. http://doi.org/10.1037/0033-

295X.106.4.766

Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world?

Behavioral and Brain Sciences , 33 , 61–83. http://doi.org/10.1017/S0140525X0999152X

Hofmann, S. G., Asnaani, A., & Hinton, D. E. (2010). Cultural aspects in social anxiety and

Social Anxiety Disorder. Depression and Anxiety , 27 , 1117–1127.

http://doi.org/10.1002/da.20759

Ioannou, K., & Fox, J. R. E. (2009). Perception of threat from emotions and its role in poor

emotional expression within eating pathology. Clinical Psychology & Psychotherapy , 16 ,

336–347. http://doi.org/10.1002/cpp.632

Jakobs, E., Manstead, A. S. R., & Fischer, A. H. (1999). Social motives and emotional feelings

as determinants of facial displays: The case of smiling. Personality and Social

Psychology Bulletin , 25 , 424–435. http://doi.org/10.1177/0146167299025004003

Joshanloo, M., & Weijers, D. (2014). Aversion to happiness across cultures: A review of where

and why people are averse to happiness. Journal of Happiness Studies , 15 , 717–735.

http://doi.org/10.1007/s10902-013-9489-9

139

Kahn, J. H., & Hessling, R. M. (2001). Measuring the tendency to conceal versus disclose

psychological distress. Journal of Social and Clinical Psychology , 20 , 41–65.

Keller, H., Lamm, B., Abels, M., Yovsi, R., Borke, J., Jensen, H., … Chaudhary, N. (2006).

Cultural models, socialization goals, and parenting ethnotheories: A multicultural

analysis. Journal of Cross-Cultural Psychology , 37 , 155–172.

http://doi.org/10.1177/0022022105284494

Keller, H., & Otto, H. (2009). The cultural socialization of emotion regulation during infancy.

Journal of Cross-Cultural Psychology , 40 , 996–1011.

http://doi.org/10.1177/0022022109348576

Kim, H. S., Sherman, D. K., Ko, D., & Taylor, S. E. (2006). Pursuit of comfort and pursuit of

harmony: Culture, relationships, and social support seeking. Personality and Social

Psychology Bulletin , 32 , 1595–1607. http://doi.org/10.1177/0146167206291991

Kitayama, S., Markus, H. R., & Kurokawa, M. (2000). Culture, emotion, and well-being: Good

feelings in Japan and the United States. Cognition & Emotion , 14 , 93–124.

http://doi.org/10.1080/026999300379003

Kitayama, S., Mesquita, B., & Karasawa, M. (2006). Cultural affordances and emotional

experience: Socially engaging and disengaging emotions in Japan and the United States.

Journal of Personality and Social Psychology , 91 , 890–903. http://doi.org/10.1037/0022-

3514.91.5.890

Kitayama, S., Park, H., Timur, A., Karasawa, M., & Uskul, A. K. (2009). A cultural task analysis

of implicit independence: Comparing North America, Western Europe, and East Asia.

140

Journal of Personality and Social Psychology , 97 , 236–255.

http://doi.org/10.1037/a0015999

Kleinknecht, R. A., Dinnel, D. L., Kleinknecht, E. E., Hiruma, N., & Harada, N. (1997). Cultural

factors in social anxiety: A comparison of social phobia symptoms and Taijin Kyofusho.

Journal of Anxiety Disorders , 11 , 157–177. http://doi.org/10.1016/S0887-

6185(97)00004-2

Knutson, B. (1996). Facial expressions of emotion influence interpersonal trait inferences.

Journal of Nonverbal Behavior , 20 , 165–182. http://doi.org/10.1007/BF02281954

Koopmann-Holm, B., & Tsai, J. L. (2014). Focusing on the negative: Cultural differences in

expressions of sympathy. Journal of Personality and Social Psychology , 107 , 1092–1115.

http://doi.org/10.1037/a0037684

Kristi Frimpong. (2013, March 28). 8 Easy Ways to Spread Happiness Around You. Retrieved

April 23, 2016, from http://tinybuddha.com/blog/8-easy-ways-to-spread-happiness-

around-you/

Lambert, N., Gwinn, A. M., Stillman, T. F., & Fincham, F. D. (2011). Feeling tired? How

sharing positive experiences can boost vitality. International Journal of Wellbeing , 1,

307–314. http://doi.org/10.5502/ijw.v1i3.1

Leu, J., Mesquita, B., Ellsworth, P. C., ZhiYong, Z., Huijuan, Y., Buchtel, E., … Masuda, T.

(2010). Situational differences in dialectical emotions: Boundary conditions in a cultural

comparison of North Americans and East Asians. Cognition and Emotion , 24 , 419–435.

http://doi.org/10.1080/02699930802650911

141

Locke, K. D. (2000). Circumplex scales of interpersonal values: Reliability, validity, and

applicability to interpersonal problems and personality disorders. Journal of Personality

Assessment , 75 , 249–267. http://doi.org/10.1207/S15327752JPA7502_6

Lu, L., & Gilmour, R. (2004). Culture and conceptions of happiness: Individual oriented and

social oriented SWB. Journal of Happiness Studies , 5, 269–291.

http://doi.org/10.1007/s10902-004-8789-5

Manser, R., Cooper, M., & Trefusis, J. (2012). Beliefs about emotions as a metacognitive

construct: Initial development of a self-report questionnaire measure and preliminary

investigation in relation to emotion regulation. Clinical Psychology & Psychotherapy , 19 ,

235–246. http://doi.org/10.1002/cpp.745

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion,

and motivation. Psychological Review , 98 , 224–253. http://doi.org/10.1037/0033-

295X.98.2.224

Matsumoto, D. (1999). Culture and self: An empirical assessment of Markus and Kitayama’s

theory of independent and interdependent self-construals. Asian Journal of Social

Psychology , 2, 289–310. http://doi.org/10.1111/1467-839X.00042

Matsumoto, D., Yoo, S. H., & Fontaine, J. (2008). Mapping Expressive Differences Around the

World The Relationship Between Emotional Display Rules and Individualism Versus

Collectivism. Journal of Cross-Cultural Psychology , 39 , 55–74.

http://doi.org/10.1177/0022022107311854

Mauss, I. B., Bunge, S. A., & Gross, J. J. (2009). Culture and automatic emotion regulation. In

M. Vandekerckhove, C. von Scheve, S. Ismer, S. Jung, & S. Kronast (Eds.), Regulating

142

Emotions (pp. 39–60). Malden, MA: Blackwell Publishing Ltd. Retrieved from

http://onlinelibrary.wiley.com/doi/10.1002/9781444301786.ch2/summary

Mauss, I. B., Butler, E. A., Roberts, N. A., & Chu, A. (2010). Emotion control values and

responding to an anger provocation in Asian-American and European-American

individuals. Cognition and Emotion , 24 , 1026–1043.

http://doi.org/10.1080/02699930903122273

Mauss, I. B., Savino, N. S., Anderson, C. L., Weisbuch, M., Tamir, M., & Laudenslager, M. L.

(2012). The pursuit of happiness can be lonely. Emotion , 12 , 908–912.

http://doi.org/10.1037/a0025299

Mauss, I. B., Tamir, M., Anderson, C. L., & Savino, N. S. (2011). Can seeking happiness make

people unhappy? Paradoxical effects of valuing happiness. Emotion , 11 , 807–815.

http://doi.org/10.1037/a0022010

McCubbin, R. A., & Sampson, M. J. (2006). The relationship between obsessive-compulsive

symptoms and appraisals of emotional states. Journal of Anxiety Disorders , 20 , 42–57.

http://doi.org/10.1016/j.janxdis.2004.11.008

McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social

networks. Annual Review of Sociology , 27 , 415–444.

McWilliams, L. A., Stewart, S. H., & MacPherson, P. S. R. (2000). Does the social concerns

component of the Anxiety Sensitivity Index belong to the domain of anxiety sensitivity or

the domain of negative evaluation sensitivity? Behaviour Research and Therapy , 38 ,

985–992. http://doi.org/10.1016/S0005-7967(99)00125-4

143

Miyamoto, Y., & Ma, X. (2011). Dampening or savoring positive emotions: A dialectical

cultural script guides emotion regulation. Emotion , 11 , 1346–1357.

http://doi.org/10.1037/a0025135

Miyamoto, Y., Uchida, Y., & Ellsworth, P. C. (2010). Culture and mixed emotions: Co-

occurrence of positive and negative emotions in Japan and the United States. Emotion ,

10 , 404–415. http://doi.org/10.1037/a0018430

Netzer, L., Van Kleef, G. A., & Tamir, M. (2015). Interpersonal instrumental emotion regulation.

Journal of Experimental Social Psychology , 58 , 124–135.

http://doi.org/10.1016/j.jesp.2015.01.006

Nils, F., & Rimé, B. (2012). Beyond the myth of venting: Social sharing modes determine the

benefits of emotional disclosure. European Journal of Social Psychology , 42 , 672–681.

http://doi.org/10.1002/ejsp.1880

Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and

collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological

Bulletin , 128 , 3–72. http://doi.org/10.1037/0033-2909.128.1.3

Parrott, W. G. (1993). Beyond hedonism: Motives for inhibiting good moods and for maintaining

bad moods. In D. M. Wegner & J. W. Pennebaker (Eds.), Handbook of Mental Control

(pp. 278–305). Englewood Cliffs, NJ, US: Prentice-Hall, Inc.

Pennebaker, J. W., Booth, R. J., & Francis, M. E. (2007). Linguistic Inquiry and Word Count

(LIWC2007).

http://homepage.psy.utexas.edu/homepage/faculty/pennebaker/reprints/LIWC2007_Oper

atorManual.pdf . Retrieved from https://researchspace.auckland.ac.nz/handle/2292/18109

144

Pharrell and “Frozen” Top iTunes’ Year-End List, Beyonce Named Best Artist. (2014,

December 8). Billboard .

Porat, R., Halperin, E., Mannheim, I., & Tamir, M. (2016). Together we cry: Social motives and

preferences for group-based sadness. Cognition and Emotion , 30 , 66–79.

http://doi.org/10.1080/02699931.2015.1039495

Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and

comparing indirect effects in multiple mediator models. Behavior Research Methods , 40 ,

879–891. http://doi.org/10.3758/BRM.40.3.879

Rimé, B., Mesquita, B., Boca, S., & Philippot, P. (1991). Beyond the emotional event: Six

studies on the social sharing of emotion. Cognition & Emotion , 5, 435–465.

http://doi.org/10.1080/02699939108411052

Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evidence for an accessibility model

of emotional self-report. Psychological Bulletin , 128 , 934–960.

http://doi.org/10.1037/0033-2909.128.6.934

Sasaki, J., & Tanno, Y. (2006). Two cognitions observed in taijin-kyofusho and social anxiety

symptoms. Psychological Reports , 98 , 395–406. http://doi.org/10.2466/pr0.98.2.395-406

Sasaki, J., Wada, K., & Tanno, Y. (2013). Understanding Egorrhea from cultural-clinical

psychology. Frontiers in Psychology , 4, 894. http://doi.org/10.3389/fpsyg.2013.00894

Sato, W., Fujimura, T., & Suzuki, N. (2008). Enhanced facial EMG activity in response to

dynamic facial expressions. International Journal of Psychophysiology , 70 , 70–74.

http://doi.org/10.1016/j.ijpsycho.2008.06.001

145

Schimmack, U., Oishi, S., & Diener, E. (2002). Cultural influences on the relation between

pleasant emotions and unpleasant emotions: Asian dialectic philosophies or

individualism-collectivism? Cognition & Emotion , 16 , 705–719.

http://doi.org/10.1080/02699930143000590

Schmidt, N. B., Mitchell, M. A., & Anthony, J. (2008). Anxiety sensitivity as an incremental

predictor of later anxiety symptoms and syndromes. Comprehensive Psychiatry , 49 , 407–

412. http://doi.org/10.1016/j.comppsych.2007.12.004

Schmidt, N. B., Zvolensky, M. J., & Maner, J. K. (2006). Anxiety sensitivity: Prospective

prediction of panic attacks and Axis I pathology. Journal of Psychiatric Research , 40 ,

691–699. http://doi.org/10.1016/j.jpsychires.2006.07.009

Shweder, R. A., Haidt, J., Horton, R., & Joseph, C. (2008). The cultural psychology of the

emotions: Ancient and renewed. In M. Lewis, J. Haviland-Jones, & L. F. Barrett (Eds.),

Handbook of Emotions (pp. 409–427). New York, NY, US: Guilford Press.

Sims, T., Tsai, J. L., Koopmann-Holm, B., & Goldstein, M. K. (2014). Choosing a physician

depends on how you want to feel: The role of ideal affect in health-related decision

making. Emotion , 14 , 187–192. http://doi.org/10.1037/a0034372

Singelis, T. M. (1994). The measurement of independent and interdependent self-construals.

Personality and Social Psychology Bulletin , 20 , 580–591.

http://doi.org/10.1177/0146167294205014

Smith, A. H., Wetterneck, C. T., Hart, J. M., Short, M. B., & Björgvinsson, T. (2012).

Differences in obsessional beliefs and emotion appraisal in obsessive compulsive

146

symptom presentation. Journal of Obsessive-Compulsive and Related Disorders , 1, 54–

61. http://doi.org/2012

Spencer-Rodgers, J., Peng, K., & Wang, L. (2010). Dialecticism and the co-occurrence of

positive and negative emotions across cultures. Journal of Cross-Cultural Psychology ,

41 , 109–115. http://doi.org/10.1177/0022022109349508

Spokas, M., Luterek, J. A., & Heimberg, R. G. (2009). Social anxiety and emotional suppression:

The mediating role of beliefs. Journal of Behavior Therapy and Experimental Psychiatry ,

40 , 283–291. http://doi.org/10.1016/j.jbtep.2008.12.004

Tamir, M., Bigman, Y. E., Rhodes, E., Salerno, J., & Schreier, J. (2015). An expectancy-value

model of emotion regulation: Implications for motivation, emotional experience, and

decision making. Emotion , 15 , 90–103. http://doi.org/10.1037/emo0000021

Tamir, M., & Ford, B. Q. (2012). When feeling bad is expected to be good: Emotion regulation

and outcome expectancies in social conflicts. Emotion , 12 , 807–816.

Tamir, M., Ford, B. Q., & Gilliam, M. (2013). Evidence for utilitarian motives in emotion

regulation. Cognition and Emotion , 27 , 483–491.

http://doi.org/10.1080/02699931.2012.715079

Tamir, M., Ford, B. Q., & Ryan, E. (2013). Nonconscious goals can shape what people want to

feel. Journal of Experimental Social Psychology , 49 , 292–297.

http://doi.org/10.1016/j.jesp.2012.11.009

Tamir, M., John, O. P., Srivastava, S., & Gross, J. J. (2007). Implicit theories of emotion:

Affective and social outcomes across a major life transition. Journal of Personality and

Social Psychology , 92 , 731–744. http://doi.org/10.1037/0022-3514.92.4.731

147

Tamir, M., Mitchell, C., & Gross, J. J. (2008). Hedonic and instrumental motives in anger

regulation. Psychological Science , 19 , 324–328. http://doi.org/10.1111/j.1467-

9280.2008.02088.x

Taylor, S. E. (2011). Social support: A review. In H. S. Friedman (Ed.), The Oxford Handbook of

Health Psychology (pp. 189–214). Oxford University Press.

Taylor, S. E., Sherman, D. K., Kim, H. S., Jarcho, J., Takagi, K., & Dunagan, M. S. (2004).

Culture and social support: Who seeks it and why? Journal of Personality and Social

Psychology , 87 , 354–362. http://doi.org/10.1037/0022-3514.87.3.354

Thoits, P. A. (2011). Mechanisms linking social ties and support to physical and mental health.

Journal of Health and Social Behavior , 52 , 145–161.

http://doi.org/10.1177/0022146510395592

Tooby, J., & Cosmides, L. (1990). On the universality of human nature and the uniqueness of the

individual: The role of genetics and adaptation. Journal of Personality , 58 , 17–67.

http://doi.org/10.1111/j.1467-6494.1990.tb00907.x

Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In J. Barkow, L.

Cosmides, & J. Tooby (Eds.), The Adapted Mind: Evolutionary Psychology and the

Generation of Culture (pp. 19–136). New York, NY, US: Oxford University Press.

Tsai, J. L., Knutson, B., & Fung, H. H. (2006). Cultural variation in affect valuation. Journal of

Personality and Social Psychology , 90 , 288–307. http://doi.org/10.1037/0022-

3514.90.2.288

148

Tsai, J. L., Louie, J. Y., Chen, E. E., & Uchida, Y. (2007). Learning what feelings to desire:

Socialization of ideal affect through children’s storybooks. Personality and Social

Psychology Bulletin , 33 , 17–30. http://doi.org/10.1177/0146167206292749

Tsai, J. L., Miao, F. F., & Seppala, E. (2007). Good feelings in Christianity and Buddhism:

Religious differences in ideal affect. Personality and Social Psychology Bulletin , 33 ,

409–421. http://doi.org/10.1177/0146167206296107

Tsai, J. L., Miao, F. F., Seppala, E., Fung, H. H., & Yeung, D. Y. (2007). Influence and

adjustment goals: Sources of cultural differences in ideal affect. Journal of Personality

and Social Psychology , 92 , 1102–1117. http://doi.org/10.1037/0022-3514.92.6.1102

Tseng, W.-S. M. D., Asaim, M. M. D., Kitanishi, K. M. D., Mclaughlin, D. G., & Kyomen, H.

M. D. (1992). Diagnostic patterns of social phobia: Comparison in Tokyo and Hawaii.

Journal of Nervous , 180 , 380–385.

Uchida, Y., & Kitayama, S. (2009). Happiness and unhappiness in east and west: Themes and

variations. Emotion , 9, 441–456. http://doi.org/10.1037/a0015634

Uchida, Y., Norasakkunkit, V., & Kitayama, S. (2004). Cultural constructions of happiness:

Theory and emprical evidence. Journal of Happiness Studies , 5, 223–239.

http://doi.org/10.1007/s10902-004-8785-9

Wang, Q. (2001). “Did you have fun?”: American and Chinese mother–child conversations

about shared emotional experiences. Cognitive Development , 16 , 693–715.

http://doi.org/10.1016/S0885-2014(01)00055-7

Wang, Q., & Fivush, R. (2005). Mother–child conversations of emotionally salient events:

Exploring the functions of emotional reminiscing in European-American and Chinese

149

families. Social Development , 14 , 473–495. http://doi.org/10.1111/j.1467-

9507.2005.00312.x

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures

of positive and negative affect: The PANAS scales. Journal of Personality and Social

Psychology , 54 , 1063–1070. http://doi.org/10.1037/0022-3514.54.6.1063

Wierzbicka, A. (1992). Talking about emotions: Semantics, culture, and cognition. Cognition

and Emotion , 6, 285–319. http://doi.org/10.1080/02699939208411073

Williams, K. E., Chambless, D. L., & Ahrens, A. (1997). Are emotions frightening? An

extension of the fear of fear construct. Behaviour Research and Therapy , 35 , 239-248.

Zaki, J., & Williams, C. (2013). Interpersonal emotion regulation. Emotion , 13 , 803–810.

http://doi.org/10.1037/a0033839

150